Blood phobia


Material Information

Blood phobia a comparison of phobics and nonphobics and an examination of affect during visual and auditory exposure
Alternate title:
Comparison of phobics and nonphobics and an examination of affect during visual and auditory exposure
Physical Description:
viii, 146 leaves : ill. ; 29 cm.
Lumley, Mark Allan, 1962-
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Phobic Disorders   ( mesh )
Blood   ( mesh )
Personality   ( mesh )
Affect   ( mesh )
Psychological Tests   ( mesh )
Adaptation, Psychological   ( mesh )
Department of Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Health Related Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1990.
Bibliography: leaves 137-144.
Statement of Responsibility:
by Mark Allan Lumley.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002357734
oclc - 50944909
notis - ALW2213
System ID:

Full Text








The doctoral dissertation documents the independent,

scholarly ability of the candidate, but it is never a

solitary effort. My efforts have been buttressed by many

others in various capacities. Foremost, I wish to note the

contributions my committee, starting with my chairperson,

Dr. Barbara Melamed, whose mentorship, support, and

friendship have made graduate school and this dissertation

pleasantly memorable. I also thank Dr. Peter Lang, whose

thoughtful critique of my research content and encouragement

to conduct a programmatic study of answerable questions have

been most helpful. My appreciation also goes to Dr. Wilse

Webb, who has gently modeled for me the role of psychologist

as discoverer and disseminator of knowledge. Dr. Sandra

Seymour is thanked for helping me keep the task in

perspective. Finally, Drs. Nancy Norvell and Anthony Greene

are acknowledged for their participation. I greatly

appreciate this committee's joy of research, desire for me

to learn, and consideration of my needs and interests.

Others have contributed in various ways. Dr.

Christopher Patrick instructed me in the conduct of

physiological assessment and the use of the data collection

software. Randle Blanco assisted in writing software

programs to convert data to analyzable format. Dr. Lars-

Goran Ost provided the surgical operations video stimulus.

Dr. Debbie Ader and Angel Siebring created the audiotapes.

Undergraduates Terry Keenan, Rick McCali, and Donna Livesey

assisted in data collection and coding. The Psychology

Department at the University of Florida provided access to

their undergraduate subject pool. The American

Psychological Association awarded a dissertation grant to

help defray costs. Finally, the National Institute of

Dental Research Training Program and the University of

Florida Presidential Graduate Research Fellowship funded my

graduate training. I am thankful for all of this aide.

Finally, to one whose sustenance I cherish, whose

caring ameliorated the tensions induced by hours in the

laboratory or at the computer, whose hope for the future

makes the trials worthwhile, I thank my wife Sherry.



ACKNOWLEDGEMENTS ....................................... ii

ABSTRACTo .................... ............. ....... vii

GENERAL INTRODUCTION................................ 1

Classification and Epidemiology.................... 1
Affective Responding in Blood Phobia................. 2
Purpose of these Studies ............................. 3

STUDY ONE ........ ... ..... ....... ..... ..... ........ 5

Introduction.... ...... ...................... ....... 5
Phobic Subjects and Controls... .............. 5
Stimulus Characteristics and Presentation Methods.. 6
Personality Dimensions............................. 8
Summary and Purpose of the Study.................. 11

Method. ....... .... . . 11
Method i... ....... . . . ........ 11
OverviewS.... .. ... .......... 11
Subjects ......h.s.... ............................. 12
Procedure ....... .. ...................... ........ 13
Stimuli...... ........................ ...... ... ... .. 17
Experimental Environment and Apparatus............. 18
Questionnaires............................ ....... 19
Dependent Measures................................. 22

Results.. .... ................................... 26
Personality Measures........................... .... 26
Videotape Stimuli................................. 27
Audiotape Stimuli .................................. .. 36

Discussion................ ....... ... .. .. .. .. 39
Psychometric Assessment............................ 39
Affect During Videotaped Stimuli.................... 41
Affect During Audiotaped Stimuli................... 47
Methodological Issues.............................. 47

Notes................ .. ....... .................... 49

STUDY TWO ........ .... ........ ...... ........ ........ .... 50

Introduction.. ..... ............. ........ ....... .. 50
Exposure and Affect Change......................... 50
Preparatory Information.............................. 51


Summary and Purpose of the Study...................

Method. ..................... ........................
Overview ................... ........................
Procedure ................. ............... ..........
Experimental Groups...............................
Stimuli ......................................... ...
Questionnaires ....................................
Dependent Measures.................................

Results ............................. ****************
Effects of Repeated Visual Exposure ................
Effects of Auditory Preparation....................
Effects of Personality Variables...................

Discussion ........... .......... ..... ** ** *
Affect Change to Repeated Surgery Exposures........
Generalization of Affect Reduction to Novel
Phobic Stimuli ...................................
Effects of Preparation ............................
Individual Differences Variables...................
Methodological Issues.............................

Notes .............. .................................

GENERAL DISCUSSION..................... ........ .......

A Comparison of Blood Phobia and Other Phobias........
Unanswered Questions...............................









IN RESEARCH (STUDY 1).................... 98



DEBRIEFING FORM (STUDY 1).................. 105



IN RESEARCH (STUDY 2)....................







DEBRIEFING FORM (STUDY 2)..................




APPENDIX K ORDER EFFECTS (STUDY 1) .................... 126


REFERENCES... ...... .................... .......... 137

BIOGRAPHICAL SKETCH...................................... 145

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




December, 1990

Chairman: Barbara G. Melamed, Ph.D.
Major Department: Clinical and Health Psychology

Two studies examined subjective, psychophysiologic

(heart rate, skin conductance level, blood pressure), and

motoric (stimulus avoidance and facial disgust expressions)

responses of blood phobics and nonphobics (defined by

elevated or below median Mutilation Questionnaire scores)

when viewing or listening to 60 s surgical and neutral

videotapes and audiotapes. Study 1 assessed several

personality domains and found that phobics (n = 24) were

more sensitive to their own anxiety, experienced greater

distress with others' negative affect, and were generally

less secure than nonphobics (n = 24). Affect was assessed

during exposure to surgical and neutral videotapes followed

by an audiotaped surgical or neutral description. Blood

phobics had more negative affect than nonphobic controls

during a surgery videotape, and phobics had greater negative

affect during a surgery than during a neutral videotape.


These differences were most prominent during only one of two

surgery scenes and when the surgery was presented prior to

rather than following the neutral videotape. Phobics and

nonphobics did not differ in affect during the neutral

scene. An audiotape describing the prior surgery elicited

slightly more arousal than a neutral description audiotape,

but phobics and controls did not differ in affect to the

description. Study 2 examined affect change to repeated

presentations of a surgery videotape, and the role of

preparatory descriptions in reducing negative affect. Sixty

blood phobics were randomly assigned to three experimental

groups. One group viewed a surgery seven times and then saw

a novel surgery; these subjects habituated during

repetitions and dishabituated to the novel surgery. Two

other groups differed in the preparation they received prior

to each of four surgery videotape repetitions. One group

heard a description of the upcoming surgery, and the second

group heard a neutral, control description. The prepared

group had moderately less negative affect during the surgery

videotapes than the control group. Individual differences

in coping style influenced responding to repeated surgery

scenes; among the prepared phobics, blunters increased

negative affect over two presentations, and monitors reduced

affect. The findings indicate the need for continued basic

research of blood phobia, especially its relationship to

fainting (which did not occur in either study) and its

current classification as a simple phobia.



Classification and Epidemiology

For some individuals, exposure to blood, bodily injury,

mutilation, disease, and related stimuli evokes a subjective

experience of fear, disgust, or aversion; autonomic arousal;

and commonly, escape from and avoidance of future encounters

with the stimulus (Marks, 1988). When this stimulus-

response pattern is sufficiently intense, it is termed

"blood phobia" (Thyer, Himle, & Curtis, 1985), and is

classified as a simple phobia in the revised third edition

of the Diagnostic and Statistical Manual of Mental Disorders

(DSM-IIIR; American Psychiatric Association, 1987).

This composite of negative subjective experience,

physiologic arousal, and escape from or avoidance of

exposure to blood and related stimuli emerges consistently

as a unitary factor in specific fear surveys (e.g., Agras,

Sylvester, & Oliveau, 1969). Additionally, fainting or

syncope during exposure to blood-related stimuli is

prevalent. Kleinknecht (1987, 1988a) found that 14.5% to

19.3% of college students reported a history of nearly or

completely fainting. About 15% of blood donors approach

syncope during or after venipuncture (Graham, 1961).

Blood phobia has been associated with reduced

motivation to donate blood, avoidance of medical or dental

visits, interference with routine tests such as

venipuncture, decreased desire for the care of one's own or

another's injuries, and the redirection of potential health

professionals from their field of interest (Lloyd & Deakin,

1975; Oswalt, 1977). Avoidance of blood-related stimuli

typically is easy for most phobics; therefore, few seek

treatment of their phobia. Nonetheless, this condition

hinders many people from full participation in activities

where blood-related stimuli occur.

Affective Responding in Blood Phobia

Emotions are best quantified by three response systems:

subjective or verbal report, physiologic activation, and

overt motor behavior (Lang, 1968). Subjectively, blood

phobics report an uncomfortable or disagreeable affect

during exposure to blood-related stimuli. Most researchers

label the emotion "fear," although no studies have

documented the occurrence of fear as opposed to a different

emotion such as disgust. Thus, this dissertation will

employ a general term such as "negative affect" to describe

the subjective component of the blood phobic's experience.

Second, some blood phobics experience a physiological

reaction, unique among the phobias, termed the "biphasic

response," in which initial sympathetic arousal is replaced

by or alternates with parasympathetic activity (Engel, 1978;

Graham, Kabler, & Lunsford, 1961). Like other simple

phobias, sympathetic activity includes tachycardia,

hypertension, striate muscle tension, perspiration, and

increased respiration, which occur during the anticipation

of or initial exposure to a blood-related stimulus. Unlike

other simple phobias, however, continued exposure may yield

parasympathetic symptoms of bradycardia, hypotension,

yawning, nausea, lightheadedness, narrowing of vision, and

ultimately fainting, if escape is precluded (Ost, Sterner, &

Lindahl, 1984). Unfortunately, many studies of blood phobia

have included only subjects who report faintness,

potentially misleading investigators to conclude that

fainting is a common, perhaps necessary concomitant of blood

phobia. Indeed, the actual prevalence of parasympathetic

symptoms and fainting per se among those who report negative

affect to and avoidance of blood-related stimuli is unknown.

Overt motor behavior is the third emotional response

domain. Like other simple phobias, blood phobics usually

physically escape from the bothersome stimulus, thus ending

the negative experience. Additionally, blood phobics appear

to successfully escape by simply turning their heads or

closing their eyes (Beck & Emery, 1985).
Purpose of these Studies

Most of our knowledge of blood phobia stems from the

treatment literature, which contains many case studies and

several controlled investigations. For example, systematic

desensitization (Babcock & Powell, 1982; Cohn, Kron, &

Brady, 1976; Elmore, Wildman, & Westefeld, 1980; Kozak &

Montgomery, 1981; McGrady & Bernal, 1986; Ost, Lindahl,

Sterner, & Jerremalm, 1984; Yule & Fernando, 1980),

implosion (McCutcheon & Adams, 1975; Ollendick & Gruen,

1972) and in vivo exposure treatments with modifications to

prevent fainting (Curtis & Thyer, 1983; Ost, Lindahl,

Sterner, & Jerremalm, 1984; Ost & Sterner, 1987; Ost,

Sterner, & Fellenius, 1989) appear efficacious in treating

blood phobia.

Although effective treatments are available, there

exists little descriptive information about the basic

psychophysiology, psychopathology, and phenomenology of

blood phobia. The process of affect change, which is

typically complicated in treatment studies, also has

received little empirical attention. This dissertation

presents two studies which attempt to increase our basic

knowledge of blood phobia. Study 1 examined differences

between blood phobics and nonphobics in their subjective,

physiological, and motoric responses to phobic and neutral

material and in several personality characteristics. Study

1 also explored differences in affect to two different

blood-related stimuli and examined the effects of stimulus

presentation order. Study 2 examined first the change in

affect during exposure to phobic material using a

habituation-dishabituation paradigm to repeated

presentations of a phobic stimulus. Study 2 also

investigated the effects of preparing subjects for exposure

to the phobic stimulus with audiotaped descriptions, and it

examined the influence of imagery ability and coping style

on affect across multiple stimulus presentations.



Phobic Subjects and Controls

Researchers typically have recruited blood phobics for

study from three sources: patients presenting for treatment

of their phobia (e.g., Ost et al., 1989), blood donaters who

faint (Graham et al., 1961), and respondents (usually

college students) with deviant scores on blood phobia

questionnaires (Beiman et al., 1978; Kleinknecht, 1988a,


Regardless of recruitment method, only a few studies

have compared blood phobics with nonphobic controls.

Klorman and colleagues (Klorman et al., 1975, 1977) found

that blood phobics (more explicitly, students with elevated

scores on the Mutilation Question [MQ], an instrument

designed to assess blood-related concerns) responded with

cardiac acceleration during 10-second exposures to

mutilation slides, whereas normals (low scoring subjects)

showed cardiac deceleration. Steptoe and Wardle (1988),

used a simple screening questionnaire (not the MQ) and found

that blood phobics reported greater anxiety and

lightheadedness and had higher heart rates and systolic

blood pressures during a surgery film than did nonphobics.

The current study also compared blood phobics (those with

elevated MQ scores) with nonphobic controls (low MQ scorers)

during exposure to phobic and neutral stimuli and on several

personality dimensions in order to enlighten fundamental

aspects of blood phobia.

Stimulus Characteristics and Presentation Methods

Research on blood phobics have employed several

different stimulus modalities. Some investigators have

assessed responding during an in vivo procedure such as

venipuncture (Engel & Romano, 1947; Graham, 1961; Kaloupek,

Scott, & Khatami, 1985), cardiac catheterization (Glick &

Yu, 1963), and pneumoencephalography (Graham, Kabler, &

Lunsford, 1961). Unfortunately, this methodology usually

lacks rigorous experimental control, exact replications are

difficult, and multiple noxious stimuli (e.g., the sight of

blood, blood loss, needles, and pain) are present. Other

investigators have used movies or films depicting surgical

procedures (Steptoe & Wardle, 1988). For example, Ost and

colleagues (Ost et al., 1989; Ost, Sterner, & Lindahl, 1984)

used a 30-minute continuous, silent videotape showing a

series of thoracic surgeries. Slides of mutilations or

homicides are the third major stimulus type employed in

blood phobia research (Hare, Wood, Britain, & Shadman, 1971;

Klorman, Weisberg, & Wiesenfeld, 1977; Klorman, Wiesenfeld,

& Austin, 1975). This methodology affords the greatest

degree of interpretive clarity, especially because the

stimulus remains static. Although a comparison has not been

done, films are expected to elicit more powerful exposure

effects than slides because of their increased similarity to

in vivo stimuli. Unfortunately, lengthy films such as that

of Ost and colleagues lead to differential viewing durations

across subjects because of fainting in some subjects.

Therefore, the current study used 60 s surgery scenes from

Ost's film. These were brief enough that all subjects were

expected to be able to watch for the full duration.

The surgical depictions in Ost's film may vary in

aversiveness for blood phobics. It is of interest to

determine the comparative aversiveness of several different

surgical scenes. Therefore, this study compared empirically

two of these surgery scenes, to determine if they elicit

different degrees of negative affect.

In addition to the above noted utility of studying

nonphobic control subjects, it is also important to compare

affect to a phobic stimulus with affect to a neutral,

nonarousing stimulus. This comparison permits conclusions

about the blood-related content of the stimulus as the

elicitor of negative affect independent of aspects of the

experimental setting involved in simply viewing a stimulus.

Steptoe and Wardle (1988) conducted such a comparison and

confirmed that blood phobics responded with less negative

affect to a neutral film than to a surgery film. This is

the only study using films for stimuli that has conducted

such a comparison, although several studies using slides

have found similar results (Hare et al., 1971; Klorman et

al., 1977, Klorman et al., 1975).

Steptoe and Wardle (1988), however, did not

counterbalance the order of surgery and neutral film

presentation, but always presented the surgery film first.

They assumed that phobics would show undesired anticipatory

anxiety to the neutral film if it were presented prior to

the surgery film. Their failure to counterbalance order

opens their findings to the alternative hypothesis that

habituation or another learning process resulted in less

negative affect during the neutral film. Study 1 corrected

for this lapse by counterbalancing stimulus order and

evaluating the effects of the two orders to determine the

validity of Steptoe and Wardle's findings.

Finally, auditory stimulus presentation is another

modality which has been employed in studies of emotion and

other anxiety disorders but not in the study of blood

phobia. Thus, Study 1 presented to both blood phobics and

controls either surgical or neutral audiotaped descriptions

of the stimulus film that they had seen earlier. Thus, the

group and stimulus content comparisons which were made for

the visual material were repeated for auditory material.

Personality Dimensions

In addition to affective differences between blood

phobics and nonphobics during surgical and neutral stimuli,

it is also of interest to study personality characteristics

of blood phobics. Study 1 examined four personality

dimensions, derived from anecdotal observations and theories

of blood phobia.

Blood phobia's etiology has received some theoretical

and empirical attention. Retrospective interviews conducted

by behavioral researchers have suggested classical and/or

vicarious conditioning etiologies similar to those found for

other simple phobias (Ost & Hugdahl, 1985), perhaps

facilitated by a genetically-based "preparedness" (Marks,

1969; Seligman, 1971). Several other hypotheses have

focused on blood phobia's uniqueness. Engel (1978)

attempted to explain both the initial anxiety and subsequent

syncope. He suggested that blood phobics, like most people,

are simultaneously sympathetically and parasympathetically

aroused by an unnatural sight such as a wound. Sympathetic

sensations consistent with the "fight or flight" response

fluctuate in dominance with feelings of queasiness and

hypotension of the parasympathetic branch. The blood phobic

is highly sensitive to his/her arousal, discomfort, and

lightheadedness, but strongly wishes to appear in control

and attempts to remain socially stoic. Thus, the phobic

feels helpless, unable either to fight or flee. The stifled

sympathetic branch is deactivated, leaving the remaining

parasympathetic activity unfettered, and syncope results.

One testable hypothesis from this theory is that blood

phobics are more sensitive to their own physical arousal

than are nonphobics. Kleinknecht (1988a) found that

subjects who reported a history of having nearly or

completely fainted had higher "anxiety sensitivity" (Reiss,

Peterson, Gursky, & McNally, 1986) than nonfainters; that

is, they were inordinately aware of, focused on, and

concerned about their physical reactions when aroused.

Study 1 attempted to replicate this finding and extend it to

blood phobics defined somewhat differently than those

studied by Kleinknecht.

Anecdotal observations suggest that blood phobics have

highly vivid visual images in response to verbal

descriptions of blood stimuli. Furthermore, blood phobics

frequently report that they "feel" in their own bodies the

injury or invasive procedure observed on another. Beck and

Emery (1985) noted that such identification with the pain or

distress of the victim induces great anxiety in blood

phobics, who experience the injury as their own. These

observations suggest that blood phobics may have greater

visual imagery abilities and a greater capacity for empathy.

Imagery ability of blood phobics has not been studied yet,

but Kleinknecht (1988a) administered a multidimensional

scale of empathy, and found that self-reported fainters had

greater feelings of personal discomfort in emotional

interpersonal situations than nonfainters. No differences

were found on other empathy dimensions such as fantasy,

perspective-taking, and general concern. Both imagery

ability and empathy were examined in this study.

Finally, a broader question is how similar blood phobia

is to other anxiety disorders or to what degree blood

phobics experience various dimensions of anxiety in their

daily life. For example, are blood phobics' muscle tension,

autonomic arousal, and feelings of fear and insecurity

greater than those of nonphobics? Or are these dimensions

no different from controls, suggesting dissimilarity to

other anxiety and stress-related disorders?

Summary of the Purpose of the Study

Study 1 attempted to increase the rigor of blood phobia

studies by including nonphobics controls and a neutral

stimulus and by counterbalancing the order of stimulus

presentation. Thus, affective differences of blood phobics

and nonphobics to surgical and neutral stimuli were

examined. Within this paradigm, the effects of stimulus

order were evaluated, as were the differences between two

surgery scenes. Additionally, audiotaped descriptions of

the surgeries were presented to determine whether the

auditory modality reliably induces affect differences across

experimental groups and stimuli. This study also examined

differences between blood phobics and nonphobic controls on

several personality dimensions of theoretical importance:

anxiety sensitivity, manifest anxiety, mental imagery, and



Forty-eight (48) volunteer undergraduate students, half

of whom were blood phobic and half of whom were nonphobic,

completed questionnaires on mental imagery, empathy, anxiety

sensitivity, and manifest anxiety. Subjects then viewed two

60-second video stimuli in counterbalanced order, including

one of two bloody surgeries and a neutral scene. Subjective

ratings, physiological responses, and facial expressions of

disgust and eye avoidance served as dependent measures.

Subsequently, half of the phobics and half of the nonphobics

heard an audiotaped description of the surgery, whereas the

remaining subjects heard a description of the neutral

videotape. Physiological and subjective measures were

assessed during the audiotapes.


Subjects were 48, 17 to 25-year-old (M = 18.9)

volunteer University of Florida undergraduates currently

enrolled in General Psychology. The final sample was

secured after screening 450 potential subjects with the

Mutilation Questionnaire (MQ) at the beginning of the

semester. Scores from the MQ were arranged in ascending

order separately for each gender. Blood phobics (n = 24)

were defined as the highest scoring 12 males and 12 females

(maximum MQ score = 30), which represented the top 6% of

each gender distribution. Scores ranged from 15 to 23 (M =

18.2) for phobic males and from 23 to 28 (M = 25.3) for

phobic females. Nonphobics (n = 24) were defined as

subjects scoring below the median of each gender

distribution. The lower half of the distribution was

divided equally into twelfths (in order to sample the full

range of nonphobics scoring below the median), and one

subject was selected from each twelfth, yielding 12

nonphobic males and 12 nonphobic females. Mutilation

Questionnaire scores ranged from 0 to 7 (M = 4.1) for

nonphobic males and from 1 to 10 (M = 5.6) for nonphobic

females. Selected subjects were contacted via telephone and

asked to participate if they were fluent in English and not

pregnant. They were told that the study involved "watching

several short television presentations and hearing several

headphone descriptions while your physical responses are

recorded." Four subjects (including one blood phobic)

declined to participate due to lack of interest and were

replaced. Subjects were paid five dollars for



The study took place during a 1 h session; subjects

were studied individually.

Psychometric assessment. After subjects read and

signed the informed consent form (see Appendix A), they

completed two randomly selected questionnaires from the pool

of four questionnaires completed during the session (see

below). Subjects then were seated in the experimental room

for physiological attachments and instructions.

Instructions. Upon entering the experimental room, the

experimenter attached to the subject an automated blood

pressure (BP) cuff and electrodes to assess heart rate (HR)

and skin conductance level (SCL). Attachments were modified

until acceptable signal quality was achieved. The BP cuff

was inflated several times prior to data collection to

accommodate the subject to its functioning. The

experimenter then read to the subject the instructions and

protocol which are presented in full in Appendix B.

Briefly, subjects were informed that they would be presented

a few short videotapes which would depict surgical and/or

neutral scenes. Following each presentation, they would

rate their emotions experienced during the presentation and

then complete a questionnaire. After several video

presentations, audiotape presentations would occur according

to the same format.

Next, the experimenter taught subjects in the use of

the Self Assessment Manikin (SAM), the computerized affect

self-report system (Appendix C). After answering subjects'

questions, the experimenter exited to the adjacent control

room, where he ascertained subjects' group status (phobic or

nonphobic), and then randomly determined the order of

stimulus presentation and which of the two surgical stimuli

were to be used.

Experimental paradigm. Table 1 presents the

experimental design and trial sequence of the study, and

Table 2 presents the timing durations and measurements for

each trial. Baseline subjective ratings were obtained prior

to the presentation of the first stimulus; subjects rated

their emotions experienced while waiting for the first

videotape presentation. After this baseline rating, the

paradigm of three trials began.

Each of the three trials followed the same data

collection format. The trial began with a baseline BP

assessment. After the cuff deflated, baseline HR and SCL

were recorded for the next 30 s, immediately prior to

stimulus onset. Contiguous with the end of this 30 s

period, the experimenter presented the stimulus to the

subject, and HR and SCL continued to be recorded during the

60 s presentation. Immediately at stimulus offset, BP was

sampled again, and the affective ratings screen illuminated

for subjects to rate the affect they experienced during the

videotape. They then waited for the next trial.

For Trials 1 and 2, one of the two surgical videotapes

(randomly selected) and the neutral videotape were

presented, counterbalancing the order of presentation both

for group and gender. For these two trials, the

experimenter presented the visual stimulus on the subject's

television, and the subject's face was videotaped during the

presentation for later analysis of avoidance behavior and

facial expression. Following the subjective ratings for

Trials 1 and 2, the experimenter reentered the experimental

chamber and gave subjects the third (after Trial 2) and the

fourth (after Trial 3) personality questionnaires;

completion time of each averaged about five minutes.

Following Trial 2 and the fourth questionnaire completion,

the experimenter placed headphones on the subject for Trial

3 (the final trial), which was a single audiotape

presentation. The experimenter randomly selected either the

surgical audiotape (describing the surgical videotape the

subject had viewed) or the neutral audiotape (describing the

Table 1. Experimental Procedure and Trial Sequence for
Study 1


# 1 & 2 Neutral # 3 Surgery # 4 Neutral

Table 2. Timing Durations and Measures for Each Trial

0 0
T Stimulus presented E
BP base IHR, SCL base HR, SCL, (Face:Trials 1,2) BP, SAM
about 30 si 30 s 60 s I about
30 s

neutral videotape the subject had viewed). The selection of

the surgical or neutral audiotape was balanced across phobia

group, gender, and the order of videotape stimulus

presentation (surgical stimulus first or neutral stimulus

first) during Trials 1 and 2. Trial 3 followed the timing

and data recording paradigm of the first two trials, except

that subjects' faces were not videotaped. After Trial 3,

the experimenter disconnected the recording devices,

debriefed subjects in accordance with APA guidelines

(Appendix D), and dismissed them.


Both phobic (surgical) and neutral videotape and

audiotape stimuli were used in this study; all stimuli were

60 s in duration. Three different videotapes were used,

including two surgical tapes and one neutral tape; all

videotapes were silent. The two surgical videotapes were

taken from the 30-minute film of thoracic operations used by

Ost and colleagues in their studies of blood phobia (e.g.,

Ost & Sterner, 1987). The two 60 s segments used in this

study depicted particularly aversive procedures in which

some cutting or piercing with a sharp instrument occurs.

"Incision" showed a scalpel incising the abdomen several

times, and other sharp instruments cutting muscle tissue.

The second surgical stimulus, "Tubes," showed a sharp tool

puncturing two holes in the abdomen and then plastic

drainage tubes being pulled through the holes. Neither

surgical scene revealed the patient's head or genitalia.

The single neutral videotape showed a wooden toy truck being

pushed over several white ramps and a person's hand picking

up and later putting down yellow blocks.

Three audiotape stimuli (two surgical and one neutral)

were employed, one corresponding to each of the three video

stimuli noted above. Each audiotape presented the voice of

a female who narrated the events in the respective videotape

in an informative, effectively neutral manner. Each

description was 60 s and 160 words long. (See Appendix E

for the transcripts of these descriptions and others used in

Study 2.)

Experimental Environment and Apparatus

Subjects were seated in a recliner with their legs

parallel to the floor, and torso reclined at approximately

30 degrees from vertical, in a 4 m X 4 m experimental

chamber. The chamber's overhead lights were off, but the

room was dimly lit by a floor lamp. A 66 cm (26 in) RCA

Lyceum color television was positioned 2 m in front of the

subject's face. This television presented the videotape

stimuli which were recorded on half-inch VHS videotapes and

played from a Panasonic videorecorder in the adjacent

control room. A 25.4 cm (10 in) Apple computer video

monitor positioned immediately to the right of the subjects'

television presented the self-report ratings display.

Subjects controlled the display by manipulating a

potentiometer knob on a control box attached to the right

arm of their chair. A black and white Panasonic videocamera

with a zoom lens was mounted on the wall in the experimental

chamber near the ceiling slightly to the left of the

television. The camera was focused on the subject's face,

permitting accurate assessment of the direction of gaze, eye

closings, and tensing of facial muscles. Videorecordings of

the subject were made on half-inch VHS videotapes in a

second Panasonic videorecorder in the control room. During

Trial 3, subjects wore comfortable Realistic NOVA 40 stereo

headphones, through which they heard the audio stimuli,

which were recorded on audiocassette tapes and played to

subjects from a General Electric stereo cassette player.

The control room was further equipped with an IBM-PC AT

which ran VPM software (Cook, Atkinson, & Lang, 1987) to

control physiological data collection and the presentation

of the affect ratings display. (See Appendix F for the VPM

control program for this study.) VPM also controlled

physiological data sampling, recording, and analysis using a

Scientific Solutions Labmaster board, an Axon Instruments

TL-1 interface panel, and Coulbourne modules. The control

room also housed a Roche Ultrasonic Blood Pressure Monitor,

Arteriosonde 1225.


Five questionnaires were used in this study.

1) Mutilation Questionnaire (MQ; Klorman, Weerts, Hastings,

Melamed, & Lang, 1974). The MQ is a 30-item, true-false

questionnaire designed to assess an individual's fear of,

discomfort with, or aversion to blood, injury, mutilation,

and related stimuli. The questionnaire's authors provided

normative data for male and female college students.

Several studies have shown the validity of the MQ relative

to psychophysiological and behavioral indices of blood-

injury concerns (Beiman et al., 1978; Green, Webster,

Beiman, Rosmarin, & Holliday, 1981; Klorman et al., 1977;

Ost, Lindahl, Sterner, & Jerremalm, 1984; Ost & Sterner,


2) Questionnaire Upon Mental Imagery (QMI; Sheehan 1967).

This questionnaire is the shortened version of Betts' (1909)

original instrument. The QMI contains 35 stimulus items

categorized in seven major sensory modalities. Subjects are

asked to rate the vividness of the images that come to mind

for each item using a seven point scale ranging from 1

("Perfectly clear and vivid") to 7 ("No image at all"). The

sum of all ratings is the total score, which ranges from 35

- 245, with low scores indicating better imagery ability.

This questionnaire has been normed (White, Ashton, & Brown,

1977) and has demonstrated reliability (Evans & Kamemoto,

1973; Hiscock, 1978) and validity (Cook, Melamed, Cuthbert,

McNeil, & Lang, 1988; Miller et al., 1987; Hiscock, 1978).

3) Interpersonal Reactivity Index (IRI; Davis, 1980). The

IRI is a 28-item, self-report questionnaire consisting of

four, factor-derived, 7-item subscales, each of which

assesses a specific aspect of empathy. The Perspective-

Taking scale measures the tendency to adopt the point of

view of other people in everyday life. The Fantasy scale

measures the tendency to transpose oneself into the feelings

and actions of fictitious characters in books, movies, and

plays. The Empathic Concern scale measures the tendency to

experience feelings of warmth, compassion, and concern for

other people. The Personal Distress scale taps one's own

emotional feelings of personal unease and discomfort in

reaction to the emotions of others. Subjects responded to

each item on a 5-point scale ranging from 0 ("does not

describe me well") to 4 ("describes me very well"). Ratings

are summed for the items in each scale, yielding four

scores. Davis (1980) reported that the four scales have

adequate internal consistency and test-retest reliability,

although females score higher than males on all scales.

Convergent and discriminant validity have been reported with

other self-report dimensions (Davis, 1983) and affective

reactions to video stimuli (Davis, Hull, Young, & Warren,


4) Anxiety Sensitivity Index (ASI, Reiss et al., 1986). The

ASI is a 12-item questionnaire that measures individual

differences in hypersensitivity to one's own anxiety

responses and behavior. Respondents endorse each item using

a 5-point scale ranging from "Very little" (0) to "Very

much" (4). A person's score on the ASI is the sum of the

scores on the 16 items. Reiss et al. (1986) provided

reliability data showing that the ASI has adequate internal

consistency and test-retest reliability. They also found

that anxiety disorder patients scored higher than non-

disordered subjects, and the questionnaire accounted for

Fear Survey Schedule variance which remained unaccounted for

by specific fear endorsement.

5) Fenz and Epstein Anxiety Questionnaire (FEQ; Fenz &

Epstein, 1965; Fenz, 1967). The FEQ is a 53-item

questionnaire listing symptoms of anxiety which are rated by

the respondent on a 5-point scale ranging from 1 ("never

applies to you") to 5 ("experience it almost all of the

time"). The scale was developed and factor analyzed to

divide manifest anxiety into its component dimensions.

Three factors are separately scored by totaling the ratings

for the items in each scale: striated muscle tension,

autonomic arousal, and feelings of fear and insecurity.

Fenz (1967) provided reliability coefficients for each scale

and found that "neurotics" scored higher than normals on all

three scales.

Dependent Measures

Dependent measures are categorized according to Lang's

(1968) three systems model of emotion: self-report indices

(i.e., verbal behavior or subjective responses),

physiological reactions, and overt motor behavior.

Self-report.1 The Self Assessment Manikin (SAM; Hodes,

Cook, & Lang, 1985; Lang, 1980) is a graphic video display

instrument for obtaining subjective ratings on three

independent affective dimensions: pleasure--displeasure,

arousal-calmness, and control--lack of control. SAM is

presented as a manikin whose features are dynamically

modifiable by subjects to represent their affect using a

potentiometer on the arm of their chair. The VPM software

program presents the three graphic displays in random order

to the subject on a video monitor. The pictorial display is

converted by the computer to a 21-point scale. In the

pleasure display, SAM's facial expression changes from a

smile to a frown; in the arousal display, SAM's "abdomen" (a

random and changing patters of dots) increases or decreases

in size and rate of change, and SAM's eyes open and close;

in the control display, SAM changes in size from very small

to very large. The validity of the three SAM ratings of

affect has been demonstrated in several studies (Cook et

al., 1988; Greenwald, Cook, & Lang, in press; Hodes et al.,

1985). For these three self-report measures, difference

scores were calculated by subtracting the baseline affect

rating (taken before Trial 1) from the affect rating for

each trial. These studies used labels for the negative

poles of each affective dimension ("displeasure," "arousal,"

"lack of control") to achieve consistency in presentation.
Physiological indices. Three different physiological

indices of affect were assessed.

1) Skin conductance level (SCL). SCL was recorded from two

Beckman Ag-AgC1 miniature electrodes placed on the thenar

and hypothenar eminence of the left hand after moistening

the palm with distilled water. A neutral paste (petroleum

jelly) was used in the electrode. Analog SCL was sampled at
10 hertz using a .5 volt constant voltage Coulbourne Skin

Conductance Coupler (S71-22), which output a digital signal

to the computer. VPM software reconverted the digital value

to micromhos, and calculated mean SCL for the 10 s baseline

immediately prior to stimulus onset and for the entire 60 s

stimulus presentation period. SCL difference scores were

calculated by subtracting the 10 s baseline SCL from the

stimulus SCL.

2) Heart rate (HR). The electrocardiogram (ECG) was

recorded from two Beckman Ag-AgC1 miniature electrodes

placed on the left and right forearms after preparing the

subject's skin with alcohol and electrode paste (Hewlett-

Packard Redux). The ECG waveform was sampled at 10 hertz

and fed from a Coulbourne bioamplifier (S75-01) into a

Coulbourne Bipolar Comparator (S21-06), which detected the

"R" wave at suprathreshold levels, and output the signal

into a Coulbourne Retriggerable One Shot (S52-12), which

then output a digital signal to the computer. VPM software

calculated R-R interbeat intervals to the nearest

millisecond and converted heart period to heart rate. Mean

HR was calculated for 10 s baseline immediately prior to

stimulus onset and for the 60 s stimulus presentation

period. HR difference scores were calculated by subtracting

the baseline HR from the stimulus HR.

3) Blood pressure. Both systolic blood pressure (SBP) and

diastolic blood pressure (DBP) were monitored using an

automated sphymomanometer with the cuff attached to

subjects' left upper arm. The experimenter manually

initiated cuff inflation from the control room and then

recorded the LED digital output by hand. Cuff inflation

required about 5 s, and deflation required about 25 s.

Difference scores were calculated by subtracting the SBP and

DBP values taken before each trial from the values obtained

at stimulus offset.

Motor behavior. Two variables were assessed from the

videotaped recordings of subjects' faces while they watched

the surgery and neutral videotapes. Two independent raters,

blind to the study hypotheses and to the videotape type

(surgery versus neutral) for each subject were trained to

code these variables.

1) Avoidance of eye contact with stimulus. This measure was

recorded as the number of seconds out of 60 that a subject's

eyes were closed or were not directed at the television

screen during the stimulus. Because the resulting

distribution was highly skewed and, therefore, not amenable

to parametric statistics, subjects were dichotomously

classified for each videotape. Avoidance was coded

positively if subjects showed at least one second of

avoidance during the videotape, and negatively if there was

less than one second of avoidance. Interrater reliability

was calculated as the percentage agreement between the two

coders. They agreed on the presence or absence of avoidance

for 47 of 48 subjects (98%) during the surgery videotape,

and they agreed on all 48 subjects (100%) for the neutral


2) Facial expressions of disgust. An evaluation of all

videotapes indicated that when subjects made a facial

expression during viewing, they routinely tensed either or

both of two muscle groups, resulting in furrowing of the

eyebrows and raising of the upper lip. According to Ekman,

Friesen, and Ellsworth (1972), this facial pattern signifies

the emotion of disgust. Thus, coders rated the maximum

degree that these muscle groups were tensed during each 60 s

videorecording using a 5-point scale (0 = "no tensing

evident," 1 "minimal tensing," 2 "mild tensing," 3 -

"moderate tensing," 4 = "severe tensing"). Since the

distribution of ratings for the nonphobics was restricted

greatly, and the vast majority of ratings for both groups

was zero during the neutral videotape, the data were treated

as frequency data and collapsed into three categories to

increase the n per cell. Original values of 0 were

classified as "none," values of 1 and 2 were classified as

"low," and values of 3 and 4 were classified as "high." The

two raters agreed on the scoring for the three levels of

facial disgust for 46 of 48 subjects (96%) during the

surgery stimulus, and 45 of 48 subjects (94%) during the

neutral stimulus.

Personality Measures

The means and standard deviations for each of the four

personality questionnaires and their subscales for the

phobics and nonphobics are presented in Table 3. Phobics

were compared with nonphobics using independent groups t-

tests. As Table 3 indicates, phobics had greater anxiety

sensitivity (ASI), feelings of personal distress (IRI-

Personal Distress), and general fear and insecurity (FEQ)

than did the controls. The groups did not differ on the IRI

subscales of Fantasy, Empathic Concern, or Perspective

Taking Ability, nor on the FEQ subscales of Muscular Tension

or Autonomic Arousal. Unexpectedly, the nonphobics reported

better mental imagery ability (QMI) than did the phobics.

Videotape Stimuli

Data analysis. A general data analytic strategy

examined affect during the two videotape stimuli (Trials 1

and 2) for displeasure, arousal, lack of control, HR, SCL,

SBP, and DBP. First, each dependent measure was analyzed in

a mixed-model repeated-measures analysis of variance (ANOVA)

in which Video (surgery and neutral videotape) was the

within-subject effect, and Group, Order (surgery videotape

presented first or neutral videotape presented first), and

Surgery ("Incision" or "Tubes" videotape) were between-

subjects variables. Significant interactions from this

model were examined via simple effects analyses using

appropriate error terms for repeated-measures models

(Howell, 1982) and corrected error degrees of freedom when

heterogeneous sources of error variance were pooled

(Satterthwaite, 1946). Typically, these simple effects

analyses examined differences between phobics and controls

for each videotape separately, and each Group was examined

Table 3. Questionnaires Scores for Phobics and Nonphobics

M (SD)

24.1 (8.2)

M (SD) t(46)

15.9 (10.4)

3.04 .004

Personal Distress
Empathic Concern
Perspective taking

Muscular Tension
Autonomic Arousal






92.0 (24.1)





74.3 (14.3)

a p-values were determined









using two-tailed tests

separately across the two videotapes. Simple effects

analyses were considered significant at the .01 probability

level. Appendix L presents the complete ANOVA tables.

Self-report measures. Displeasure, arousal, and lack

of control change scores from baseline for the phobics and

nonphobics for both surgery and neutral videotapes are

presented in Table 4. Similar results were found for all

three measures. During the surgery videotape, phobics

reported more arousal, displeasure, and lack of control than

did the nonphobics, whereas during the neutral videotape,

the two groups did not differ in any measure. Across

videotapes, the phobics reported greater negative affect on

all three measures to the surgery than to the neutral

videotape. The nonphobics reported greater arousal to the

surgery than to the neutral videotape, but no difference in

control to the two videotapes, (Video X Group interactions

for displeasure, E(1, 40) = 34.91, R < .0001; arousal, F(1,

40) = 14.76, R < .0004; and lack of control, F(l, 40) =

24.33, p < .0001). For displeasure, however, there was an

influence of videotape order (see Appendix K for these

data). Phobics reported more displeasure to the surgery

videotape when it was presented before, but not after, the

neutral videotape. Nonphobics reported more displeasure to

the surgery videotape than to the neutral videotape only

when the surgery was presented after, but not before, the

neutral videotape, (displeasure Video X Group X Order

interaction, f(l, 40) = 8.73, R < .006). Videotape order

Table 4. Self-Reported Changes in Displeasure, Arousal,
and Lack of Control During the Surgery and
Neutral Videotapes for Phobics and Nonphobics

Videotape Stimulus

Surgery Neutral
M (SD) M (SD)
Displeasure 9.7 (4.6) -0.7 (2.6)
Arousal 8.1 (6.9) -4.0 (6.7)
Lack of Control 4.7 (6.3) -5.4 (4.4)

Displeasure 2.8 (3.8) -0.3 (3.2)
Arousal 2.8 (5.7) -3.6 (4.2)
Lack of Control -1.9 (5.0) -4.1 (4.5)

did not affect arousal or control, and surgery type ("Tubes"

or "Incision") was unrelated to any self-report variable.

Physiological measures. Figure 1 presents mean SCL and

HR change scores during the "Incision" and "Tubes" surgery

videotapes and the neutral videotape for phobics and

nonphobics. As the figure reveals, the SCL and HR of the

phobics who viewed "Incision" was greater than the SCL and

HR of a) nonphobics who viewed "Incision," and b) phobics

who viewed "Tubes." Phobics and nonphobics did not differ

in SCL or HR during "Tubes," nor did the nonphobics' SCL or

HR differ to the two surgeries. Across videotapes, both

phobics and nonphobics had a greater SCL during the surgery

than the neutral videotape, regardless of Surgery type.

However, HR change was greater during the surgery than the

neutral videotape only for phobics who viewed "Incision,"

but not for phobics who viewed "Tubes" or for the

nonphobics. Phobics and nonphobics did not differ in SCL or

HR during the neutral videotape, (for SCL: Group X Surgery,

f(1, 40) = 4.30, E = .044, and Video X Group X Surgery, E(l,

40) = 3.74, R = .06; for HR: Video X Group X Surgery, F(l,

40) = 12.07, R = .001).

Although videotape Order failed to significantly affect

SCL, it did influence HR (see Appendix K). Phobics who

viewed a surgery before viewing the neutral videotape had a

greater HR increase than a) nonphobics who saw a surgery

first, and b) phobics who saw a surgery second. Across

videotapes, only those phobics who saw a surgery before the

PI.....PI Phobics, Incision
PT---PT Phobics, Tubes
NI- -NI Nonphobics, Incision
NT._._.NT Nonphobics, Tubes

2.0 +

6 +









4 +

PT \


\ .\ .*
\ \ .



0 +




-2 H



-4 +


Figure 1.

Skin Conductance Level Change and Heart Rate
Change Across Surgery and Neutral Videotapes for
Phobics and Nonphobics by Type of Surgery Viewed

1.5 +











neutral videotape had a greater HR during the surgery than

during the neutral, (Video X Order, F(1, 40) = 12.79, R =

.0009; Group X Order, f(1, 40) = 4.16, R = .048).

Analyses of changes in SBP and DBP failed to reveal any

significant effects or even nonsignificant trends within or

across videotapes or as a function of Group, Order, or

Surgery for either dependent measure.

Motor behavior measures. Avoidance and the maximum

facial expression of disgust during the videotapes were

analyzed as frequency data using chi-square analyses.

Tables 5 and 6 present these data for both phobics and

nonphobics for both videotapes by type of surgery and order

of presentation. Phobics more frequently avoided and showed

"high" disgust2 than did the nonphobics during the surgery

"Incision" but not during "Tubes," (avoidance, X2(1) = 8.71,

R < .005; disgust, X2(1) = 9.88, R < .005). Among the
phobics, avoidance was significantly more frequent, and

"high" disgust tended to be more frequent during "Incision"

than during "Tubes," (avoidance, X2(1) = 6.171, R < .013;

and disgust, X2(1) = 2.74, R < .10). Avoidance and disgust

of the nonphobics was not influenced by the type of surgery.

With respect to stimulus order, phobics avoided the surgery

videotape more than the nonphobics only when the surgery was

shown first, X2(1) = 5.04, R < .025; but the groups did not

differ in avoidance of the surgery when it was presented

after the neutral videotape. Stimulus order did not affect

the phobics or nonphobics' avoidance of the surgery when

Table 5. Number of Phobics and Nonphobics Displaying
Avoidance Behavior During the Surgery and Neutral
Videotapes by Type of Surgery Viewed and Order of
Videotape Presentation

Videotape Stimulus
Surgery Neutral

Yes No Yes No

Phobics (n=24) 10 14 4 20

Type of Surgery
"Incision" (n=12) 8 4 4 8
"Tubes" (n=12) 2 10 0 12

Order of Presentation
Surgery first (n=12) 6 6 3 9
Neutral first (n=12) 4 8 1 11

Nonphobics (n=24) 2 22 2 22

Type of Surgery
"Incision" (n=12) 1 11 0 12
"Tubes" (n=12) 1 11 2 10

Order of Presentation
Surgery first (n=12) 1 11 1 11
Neutral first (n=12) 1 11 1 11

Table 6. Number of Phobics and Nonphobics Displaying Facial
Expressions of Disgust During the Surgery and
Neutral Videotapes by Type of Surgery Viewed and
Order of Videotape Presentation


none low

Phobics (n=24)

Type of Surgery


Disgust Level
high none low

5 9 10

2 0


Order of Presentation
Surgery first (n=12)
Neutral first (n=12)

Nonphobics (n=24)

8 16 0

3 0

Type of Surgery


Order of Presentation
Surgery first (n=12) 5 7 0 10 2 0
Neutral first (n=12) 3 9 0 11 1 0


each group was examined alone. The disgust of both groups

was not affected by the order of presentation. Finally,

during the neutral videotape, phobics did not differ from

nonphobics in avoidance or disgust.

Next, the frequency of avoidance and disgust across the

two videotapes was examined.3 Significantly more phobics (7

of 8) differentially avoided the surgery rather than the

neutral videotape, X2(1) = 4.50, E < .035; and more phobics

(19 of 19) showed higher levels of disgust during the

surgery videotape than during the neutral videotape, X2(1) =

19.0, R < .0001. The nonphobics did not differentially

avoid the two stimuli, but all 13 of the nonphobics who

differentially showed disgust to the two videotapes showed

higher disgust levels during the surgery videotape, X2(1) =

13.0, E < .001. Surgery type and presentation order did not

influence disgust across videotapes.

Audiotape Stimuli

Data analyses. Dependent measures assessed during the

audiotape presentation (Trial 3) included displeasure,

arousal, lack of control, HR, SCL, SBP, and DBP. Motor

behavior was not assessed during the audiotape trial. Data

analyses for these dependent measures used univariate ANOVAs

which included the following between-subjects effects and

their interactions: Audio (whether the subject heard the

surgery or the neutral audiotape), Group, Order (of the

preceding surgery and neutral videotapes), and Surgery.

Self-report measures. Data for the three self-report

measures for phobics and nonphobics by type of audiotape

presentation (surgery or neutral) are presented in Table 7.

The results for displeasure and control lack are identical.

Phobics who heard a surgery audiotape reported more

displeasure and lack of control than nonphobics who heard a

surgery audiotape, and more displeasure and lack of control

than phobics who heard the neutral audiotape. These two

variables did not differ for the nonphobics during the two

audiotapes. The two groups did not differ in displeasure or

lack of control during the neutral audiotape, (Group X Audio

interactions for displeasure, F(l, 32) = 7.95, R < .009; and

lack of control, F(l, 32) = 10.57, R < .003). For arousal,

whether subjects were phobic or not, greater arousal was

reported to the surgery audiotape than to the neutral

audiotape, (Audio main effect, F(1, 32) = 27.03, E < .0001).

For all three self-report measures, there was no difference

in response to "Incision" or "Tubes" or to the order in

which the subjects had viewed the preceding videotapes.

Physiological measures. Subjects who heard the surgery

audiotape had a higher SCL (M = -0.21, SD = 0.56) than

subjects who heard the neutral audiotape (M = -0.51, SD =

0.36), Audio main effect, F(l, 32) = 5.08, R < .032. There

were no differences between phobics and nonphobics, nor were

there differences as a function of Surgery or Order. For

HR, SBP, and DBP, the ANOVAs revealed no significant effects

of Audiotape, Group, or the other independent variables.

Table 7. Self-reported Change in Displeasure, Arousal,
and Lack of Control During the Audiotape
Presentation for Phobics and Nonphobics by Type of
Audiotape (Surgery or Neutral)

Audiotape Stimulus

Surgery Neutral
Phobics M (SD) M (SD)

Displeasure 7.7 (4.7) 0.3 (3.6)
Arousal 4.8 (4.6) -2.6 (7.3)
Lack of Control 3.1 (4.9) -6.3 (3.5)


Displeasure 2.7 (2.5) 1.9 (2.8)
Arousal 3.2 (4.9) -5.4 (5.2)
Lack of Control -1.7 (3.8) -3.1 (4.3)


This study accomplished several goals. First, it

compared blood phobics with nonphobic controls on specific

personality dimensions and found differences of theoretical

importance. Second, affect (subjective, psychophysiologic,

and motor) of phobics and nonphobics was examined during

surgical and neutral videotapes. Phobics displayed more

negative affect to the surgical presentation than did the

nonphobics, but the groups did not differ during the neutral

stimulus. Both groups had greater negative affect to the

surgical than to the neutral stimulus. Third, two different

surgery videotapes were compared, and one was more aversive

than the other. Fourth, the order of surgical and neutral

stimulus presentations was evaluated and found to have some

effect on responding. Finally, phobics and nonphobics heard

surgical or neutral audiotape presentations, and the

observed differences were fairly limited.

Psychometric Assessment

Theory and anecdotal observations specific to blood

phobia guided the assessment of four personality dimensions.

Consistent with the findings of Kleinknecht (1988a), blood

phobics reported greater "anxiety sensitivity," suggesting

they are more attuned to and concerned about physiological

indices of their own arousal than are nonphobics.

Unfortunately, the questionnaire assessed sympathetic

aspects of anxiety and not the parasympathetic activity

which blood phobics may experience. Nonetheless, if one can

extrapolate from the current assessment of anxiety to

parasympathetic symptoms, then this finding supports Engel's

(1978) hypothesis that blood phobics are hypervigilant about

and overly fearful of their own negative physical and

emotional responses to blood-related stimuli. Future

research should assess parasympathetic symptoms to test this


The Fenz-Epstein Anxiety Questionnaire assessed the

degree to which blood phobics differ from normals and are

like other anxiety and psychosomatic disordered patients

("neurotics"). Blood phobics did not have elevated general
levels of autonomic arousal or muscular tension, suggesting

that they differ from the classic neurotic pattern of

anxious tension and autonomic liability. However, like other

"neurotic" patients, blood phobics reported greater general

fear and insecurity than controls. This finding is

consistent with Engel's (1978) model in that blood phobics

may overly seek to appear socially adequate and in control,

especially when experiencing discomforting physical and

emotional sensations.

The suggestion that blood phobics strongly identify

with the victim (Beck & Emery, 1986)--that they are able to

"feel" what the other person is feeling or to "get under the

skin" of the other--was assessed via questionnaires of

empathy and mental imagery. On the empathy scale, blood

phobics and controls did not differ in their ability to

fantasize, to take another's perspective, or in their

concern and caring about others. Yet blood phobics rated

themselves as more likely to feel personal distress or

unease in reaction to other's negative emotions. This

finding replicates that of Kleinknecht (1988a) and helps to

operationalize the notion of identifying with a victim.

Thus, an injury or pain in another person precipitates quite

readily negative affect in the blood phobic.

Mental imagery ability was assessed also to examine

identification with the victim. Unexpectedly, nonphobic

controls reported more vivid mental imagery than phobics.

This may be due not to poor imagery of phobics, but rather

to a nonphobic sample with unusually good imagery scores.

Replication with another sample is needed to confirm the

group difference.

It is tempting to speculate that the increased anxiety

sensitivity, general fear and insecurity, and personalized

empathic distress predated and abetted the full-blown

phobia; however, it is quite possible that these

characteristics resulted from one or more negative reactions

to blood-relevant stimuli. Longitudinal research is needed

to clarify this interpretive bind.

Affect During Videotaped Stimuli

This study used neutral stimulus material to control

for the effects of the experimental setting and stimulus

viewing and to permit definitive conclusions regarding the

affective responding of phobics and nonphobics to phobic

material. Thus, all subjects viewed a 60 s neutral

videotape in addition to a 60 s surgical videotape.

Equivalent levels of negative affect were displayed by both

groups during the neutral videotape, suggesting that any

observed differences during the surgical scene were not

attributable to extraneous factors of the experimental


Given the affective equivalence of phobics and

nonphobics to the neutral stimulus, it was expected that

phobics would report greater negative affect, be more

physiologically aroused, and show more avoidance and facial

expressions of disgust during a surgical videotape than

would nonphobics. Indeed, for most dependent measures, the

average scores of the blood phobics as a group indicated

that they had greater negative affect. Yet, the observed

differences for several dependent measures were limited by

the particular surgery scene viewed and/or the order of

videotape presentation.

For HR, SCL, facial disgust, and avoidance, differences

between phobics and controls were limited to the surgery

videotape, "Incision," which depicted a scalpel cutting a

person's abdomen and sharp tools opening the wound. Phobics

who viewed this surgery responded more negatively than did

nonphobics who viewed "Incision." No differences were found

between phobics and nonphobics during the surgery videotape

"Tubes," which showed a pliers-type tool puncturing a

person's abdomen and pulling plastic drainage tubes through

the holes. The nonphobics responded with a fairly low level

of negative affect to both surgeries; however, among the

phobics, "Incision" was more aversive than "Tubes." The

observed differences between the two surgical scenes is

interesting, in that both scenes show blood and the

"mutilation" of a portion of one's abdomen.

There are several possible explanations for the

observed differences. First, although the two videotapes

were similar in gross aspects, "Incision" simply may be a

more powerful aversive stimulus than "Tubes," in that a

scalpel cutting may be more aversive than a pair of pliers

puncturing and pulling. A second, perhaps related

hypothesis pertains to the observers' comprehension of the

events depicted. In this study, subjects had been informed

previously only that the videotapes showed portions of a

surgery on a living human being. The insertion of the tubes

might have not been recognized readily as a surgical

procedure, in contrast to the clearly recognized scalpel

incision. Thus, "Tubes" might have evoked increased

curiosity, and hence, less aversion than "Incision." This

hypothesis might be empirically examined by providing

subjects a description of the "Tubes" surgery before or

during viewing, thus eliminating uncertainty about content.

Affective group differences to the surgery videotapes

were limited also by the order of videotape presentation.

During the surgery presentation, phobics had higher HRs and

showed more avoidance than controls only when the surgical

stimulus was presented first rather than second; when the

surgery was presented after the neutral, the groups did not

differ. Among phobics, a surgical scene viewed first

elicited more displeasure as well as tachycardia and

avoidance than did a surgical presentation after the neutral


A parsimonious explanation for the observed effects is

that habituation occurred, and subjects felt generally less

aroused to a later presentation of a surgery than to an

earlier presentation due to the passage of time and

accommodation to the experimental setting. An alternative

explanation is that subjects acquired information during the

earlier presentation of the neutral stimulus which resulted

in lower anxiety on a subsequent presentation. Information

potentially acquired during the initial neutral exposure

included the duration of the stimulus, the functioning of

the television, and a reduction in ambiguity regarding which

stimulus (surgery or neutral) they were most likely to see

next. A test of these two hypotheses might be achieved by

comparing two groups of phobics, one which views a neutral

followed by a surgery film, whereas the other waits an equal

length of time before viewing a surgery film.

Alternatively, order effects might be controlled by

providing a practice videotape trial before presenting the

two stimuli of interest and informing subjects of the order

of scenes.

Thus, several variables other than the presence or

absence of blood phobia influenced responding to surgical

stimuli. It should be remembered, however, that the

observed effects of surgery type and/or order may be due to

bias in random assignment; for example, the 12 phobics

assigned to view "Incision" or to view a surgery before the

neutral videotape may have differed in some important, yet

unassessed way from other phobics. The need for replication

on a different and larger sample of phobics is clear.

In addition to comparisons between phobics and

nonphobics, this study examined each group's affect to both

surgical and neutral stimuli. As expected, phobics reported

greater displeasure, lack of control, and arousal during the

surgery videotape than during the neutral videotape.

Nonphobics also reported greater arousal to the surgery

stimulus, but limited or absent differences in displeasure

and control. On physiological measures, the differences

between videotapes were less robust. Both phobics and

nonphobics had a higher SCL during the surgery videotape in

comparison to the neutral videotape. For HR, only those

phobics who viewed "Incision" or who viewed a surgery first

were more tachycardic during the surgery than during the

neutral videotape. The HR of other phobic subgroups and of

the nonphobics did not differ during the two stimuli. Blood

pressure measures did not differentiate the two videotapes

for either group. Finally, both phobics and nonphobics more

frequently displayed facial expressions of disgust to the

surgery than to the neutral videotape; however, only the

phobics more frequently avoided the surgery than the neutral

videotape. In summary, phobics clearly showed more negative

affect during the surgery videotape than during the neutral

videotape, especially for the surgery "Incision," whereas

nonphobics showed similar but less pronounced and less

consistent affective differences to the two stimuli. The

observation of some increase in negative affect to the

surgery scene for the nonphobics is consistent with the view

that aversiveness to blood-related stimuli exists on a

continuum. Whereas extreme cases may be considered phobic,

cases selected from other portions of the distribution (such

as below the median) have relatively less aversion, but

still more than to a neutral stimulus.

Steptoe and Wardle (1988) suspected that when both

blood-related and neutral material were presented to blood

phobics, the order of presentation would be important. In

their study, they eschewed stimulus counterbalancing and

presented to all subjects the bloody stimulus first followed

by the neutral stimulus. They presumed that the

presentation of the neutral stimulus first to half of the

phobics would have yielded a biased sample of "neutral"

responding, with elevated fear due to anticipatory anxiety

over the upcoming phobic stimulus. The current study tested

their assumption and found, contrary to their hypothesis, no

evidence that responses during a first presentation of the

neutral videotape differed from responses during the second

neutral stimulus presentation. Furthermore, the order

effects found in this study suggest that Steptoe and

Wardle's presentation of the blood stimulus first probably

resulted in greater negative affect to it, and greater

differences between the blood and subsequent neutral


Affect During Audiotaped Stimuli

This study also assessed affective responding during

verbal descriptions of phobic and neutral stimuli in order

to understand how various stimulus presentation modalities

influence the affect of blood phobics. All subjects heard

an audiotaped description of either the surgery or the

neutral scene that they had just seen.

Phobics who listened to a surgical description reported

more displeasure and control lack than phobics who heard the

neutral description or nonphobics who heard a surgical

description; however, only these two variables discriminated

conditions. Self-reported arousal and SCL were increased

during the surgery audiotape regardless of whether a person

was phobic or not, and neither HR nor BP measures differed

as a function of stimulus content or group. In summary, the

effects of group and stimulus type during audiotape

presentations were less robust than during videotape

presentations. Potential reasons are discussed later.

Methodological Issues

It must be acknowledged that the observed differences

between phobics and nonphobics and also between surgical and

neutral stimuli were of limited magnitude. First, the

subject selection procedure probably reduced group

differences. Phobics and nonphobics were selected from

different points of a nearly normal distribution of MQ

scores using rather arbitrary cut-scores. Restriction of

those ranges by studying only the few highest scoring

phobics or lowest scoring nonphobics might have increased

the observed effect size. Alternatively, studying phobics

presenting for treatment (although such people are rare)

likely would have yielded more clear differences.

Limited differences also were found between surgical

and neutral stimuli. In addition to the effects of stimulus

presentation order and surgery type described above, the

type of phobic stimulus undoubtedly had an impact. First,

it is possible that a videotaped portion of a surgery is

less aversive than, for example, mutilation scenes such as

injuries, accidents, etc. Additionally, although the two

surgery scenes were taken from an apparently powerful phobic

stimulus of thoracic operations (Ost, Sterner, & Lindahl,

1984), the presentation methods differed markedly. Ost and

colleagues showed a continuous 30 minute videotape that

included scenes of the full human patient during surgery.

In the current study, the elimination of scenes showing the

patient's face or full body might have reduced the negative

impact of the stimulus. Indeed, during debriefing, many

phobics noted that the operation seemed somewhat unreal,

partly because they did not know for sure that a human was

undergoing surgery. Finally, unlike the methodology of Ost

and colleagues in which the 30 minute stimulus was not

stopped until fainting or continuous avoidance occurred,

this study's use of a 60 s stimulus permitted the collection

of uniform data for all subjects but probably also decreased

exposure intensity. In summary, a more intense phobic

stimulus probably would have increased group differences.

1 An attempt was made to assess feelings of faintness and of
nausea independent from the three SAM scores by presenting
two VPM-generated visual analog scales to subjects following
each of the SAM ratings. Scores from these faintness and
nausea scales were found to be highly correlated with the
self-reported displeasure (r = .66 and .70, respectively),
to be much more highly endorsed by females (unlike the three
SAM measures), and to be highly skewed with most scores
being zero. Additionally, given the experimental nature of
these scales and the lack of prior psychometric data, these
scales are not presented in this manuscript.
2 When examining the effects of Surgery and Order on facial
disgust, three levels of disgust would have resulted in
unacceptably low sample sizes for chi-square analyses. Thus,
for these analyses, a single "low" category was created by
collapsing the "none" and "low" categories.

3 Chi-square analyses across stimuli (within-subjects) must
not violate the assumption of independent observations;
therefore, these analyses for avoidance and facial disgust
utilized McNemar's (1969) suggestion to determine if the
number of subjects who, for example, showed avoidance
differentially during the two videotapes varied
significantly from expected. The null hypothesis in this
case is that the number of subjects who avoided the surgical
videotape but not the neutral videotape equals the number of
subjects who avoided the neutral videotape but not the
surgery. Chi-square goodness of fit tests evaluated this

4 The small number of subjects who showed differential
avoidance to the two videotapes precluded an analysis of the
effects of surgery type and presentation order.



The goals of Study 2 were to a) evaluate the presence

and extent of habituation to repeated exposures to a

surgical visual stimulus and dishabituation to a novel

surgical stimulus; b) assess the effect on habituation and

dishabituation of preparing blood phobics by providing

either a relevant description of the upcoming surgery or a

neutral control description; and c) examine how imagery

ability and coping style moderate the effects of preparation

and repeated exposures.

Exposure and Affect Change

A wealth of clinical data strongly supports the

proposal that exposure--by various means--changes affect

(Barlow, 1988; Foa & Kozak, 1985; Marks, 1978). In

particular, these researchers agree that exposure to

anxiety-evoking stimuli produces the changes noted during

treatment of anxiety disorders and phobias. It is

noteworthy that each of the therapeutic interventions for

blood phobia noted in the General Introduction incorporates

some form of exposure to blood-relevant material.

Yet few studies have examined the process of emotional

change during exposure to aversive stimuli in blood phobia.

Hare et al. (1971) studied normal adults and found that

repeated presentations of the same mutilation slide resulted

in rapid physiological habituation, whereas different

mutilation slides interfered with habituation. Similar

studies have presented mutilation slides with the purpose of

studying orienting and defense reactions in blood phobics

(Klorman et al., 1975, 1977). The use of repeated exposure

to videotapes or films is rare (see Klorman, 1974 for an

exception) and has not been conducted with blood phobics.

It is difficult to achieve prolonged exposure in blood

phobia without incorporating syncope-preventing strategies

or suffering missing data. Therefore, the current paradigm

used brief, repeated exposures to surgical stimuli to

produce habituation in blood phobics.

Studies of repeated exposures to aversive stimuli

typically have not examined the generalization of reduced

affect to related stimuli once extinction or habituation to

one stimulus has occurred. Yet repeated visual exposures to

a blood-related stimulus may result in reduced negative

affect to a novel blood-related stimulus secondary to

modifications in the "blood-related" emotional network.

Preparatory Information

Exposing subjects to films was the paradigmatic

approach of Lazarus and colleagues, who explored the

efficacy of verbal "defensive sets" to modify stress

responses during mutilation and other stressful films

(Lazarus & Alfert, 1964; Lazarus, Speisman, Mordkoff, &

Davison, 1962; Speisman, Lazarus, Mordkoff, & Davison,

1964). In these studies, introductory statements or film

soundtracks were modified to induce "intellectualization" or

"denial" sets. The finding of group differences on some

physiological measures prompted the authors' claim that

defensive sets effectively "short-circuited" stress.

Alternatively, the presentation of accurate preparatory

information may have led to decreased physiological arousal.

A large literature indicates the need to consider the

affect-modifying role of preparatory information. Many

studies have demonstrated that preparing medical and dental

patients with information about upcoming aversive procedures

reduces physiological arousal, behavioral escape or

avoidance, and subjective distress at various points before,

during, and after the procedure (Anderson & Masur, 1983;

MacDonald & Kuiper, 1983; Rogers & Reich, 1986; Silver &

Wortman, 1980). These preparation researchers have not

discussed the affect-modifying effects of information in an

emotional imagery framework (Hebb, 1968; Lang, 1979, 1985).

In this view, a perceptual-motor memory is activated,

observable visceral and motor responses occur, and

modification of the memory's stimulus, meaning, and response

propositions is facilitated. Thus, preparatory information

may reduce fear of an event by activating the relevant

emotional network and permitting fear modification.

Several individual difference variables may influence

the effect of preparatory information and/or the change in

affect across exposure repetitions. First, individuals

differ in their ability to create vivid images from verbal

prompts, with resulting differences in visceral activity,

and these imagery ability differences can be reliably

assessed via questionnaire (Miller et al., 1987). Good

imagers processing fear-relevant verbal descriptions

demonstrate more arousal than do poor imagers (Cook et al.,


Second, a person's coping style may influence

preparation effects and emotion during single or repeated

presentations of an aversive stimulus. Briefly, the coping

style literature claims that people consistently differ in

their preferred manner of dealing with aversive stimuli.

Some people preferentially gather information about and

directly engage the stimulus in order to decrease negative

affect. Others tend to avoid information about the stimulus

and negate its relevance and impact. These two coping

styles variously have been termed sensitization and

repression (Byrne, 1961), confrontation and avoidance (Suls

& Fletcher, 1986), and monitoring and blunting (Miller,

1980), and can be assessed reliably via questionnaires such

as Miller's (1980) instrument. Prior research on preparing

medical patients for noxious procedures found that

continuous presentations of information decreased anxiety of

monitors but increased anxiety of blunters (Shipley, Butt, &

Horwitz, 1979; Shipley, Butt, Horwitz, & Farby, 1978). Of

importance to the current study, information and stimulus

repetition may interact with subjects' coping style,

yielding different patterns of affect change across


Summary and Purpose of the Study

This study addressed several questions about exposure

and affect change in blood phobia. First, does repeated

exposure to a phobic stimulus (surgery videotape) result in

decreased negative affect over repetitions? If such

habituation occurs, to what degree does it generalize to a

novel blood-related stimulus? These questions were

addressed by exposing one group of blood phobics (Video Only

Group) to seven repetitions of a surgical videotape followed

by one novel surgery presentation.

Second, this study examined the effects on habituation

and dishabituation of providing blood phobics with

preparatory verbal descriptions before each surgery

videotape exposure. Two types of verbal descriptions were

used. One group of phobics (Surgery Audiotape Group) heard

a factual description of the upcoming surgery before each

repetition. They were compared with a control group of

phobics (Neutral Audiotape Group) who heard an irrelevant

preparatory description before each surgery videotape

repetition. It was expected that the prepared phobics would

have less negative affect than the controls during each

surgery videotape. Finally, this study examined the effects

of imagery ability on the affect of both groups during the

preparatory descriptions, and it examined the effects of

coping style on affect during surgery videotape exposures.


Sixty volunteer adult blood phobics were interviewed,

completed questionnaires of imagery ability and coping

style, and were randomly assigned to one of three groups:

Video Only, Surgery Audiotape, or Neutral Audiotape. Video

Only subjects were exposed seven times to a surgical

videotape, followed by a single exposure to a novel surgery.

Subjects in the Surgery Audiotape and Neutral Audiotape

Preparation Groups experienced a different exposure

paradigm. Four presentations of a preparatory audiotape

alternated with four presentations ("test trials") of a

surgery videotape. Subsequently, these subjects viewed a

novel surgery videotape. These two groups differed only in

the type of preparatory audiotape they heard prior to each

of the four surgery test trials: Surgery Audiotape subjects

heard a factual description of the upcoming surgery, and

Neutral Audiotape subjects heard a neutral irrelevant

description. Subjective affect, physiological responses,

and facial expressions of disgust and avoidance served as

dependent measures for all three groups.


Subjects were 60, 18 to 25-year-old (M = 19.7)

volunteer University of Florida undergraduates currently

enrolled in General Psychology. The final sample was

secured after screening approximately 500 potential subjects

with the Mutilation Questionnaire (MQ) at the beginning of

the semester. The highest scoring males and females were

telephoned and invited to participate in a study if they

spoke English and were not pregnant. One male declined to

participate fearing his anticipated reaction to the stimuli,

and one female with panic disorder was excluded. Mutilation

Questionnaire scores of blood phobic males (n = 30) ranged

from 14 to 24 (M = 17.4), and scores of blood phobic females

(n = 30) ranged from 21 to 29 (M = 23.6). Approximately the

upper 15th percentile of each gender distribution was used

from this sample. All subjects received course credit for

participating in the experiment.


Each subject was studied individually during a 2 h

experimental session.

Interview and psychometric assessment. After subjects

read and signed the informed consent form (Appendix G), they

completed severally randomly ordered questionnaires,

including the Questionnaire Upon Mental Imagery and the

Miller Behavioral Style Survey.

Experimental session. Following questionnaire

completion, subjects entered the experimental room,

instructions were presented (Appendix H), and the electrodes

for HR and SCL and the BP cuff were attached. The cuff was

inflated several times to accommodate the subject to its

function. Next, all subjects had one practice videotape

trial using a neutral video stimulus in order to accommodate

them to the videotape presentation. The protocol for this

practice trial was the same as for all other trials

described below, except that no data were collected during

this practice trial. Following the practice, the

experimenter reentered the room, answered the subject's

questions, placed the audio headphones on the subject, and

departed. In order to accommodate subjects to the

headphones and the speaker's voice, subjects heard over the

headphones a brief reminder about the instructions of the

study (Appendix H). Following these headphone reminders,

subjects rated the affect they experienced during the

headphone instructions. These ratings served as the study

baseline subjective ratings. Subjects then waited several

minutes for the first presentation, during which the BP cuff

was inflated twice, one minute apart, and the average of

these two BP assessments served as the study baseline BP.

Group and stimulus assignment. Prior to the start of

the study, a research assistant randomly assigned subjects

to one of the three experimental groups (in blocks of three

or six subjects) and matched the groups for gender,

resulting in 20 blood phobic subjects--10 males and 10

females--per experimental group. The group assignment for

each subject was placed in an envelope and opened by the

experimenter after his last interaction with the subject

prior to the start of the trials, thus keeping the

experimenter blind to group assignment during his

interactions with the subject. The assignment of the

particular stimulus (or stimulus pair for the Neutral

Audiotape group) for each subject also was determined

randomly prior to the study using Latin squares. Stimulus

assignment was conducted to assure that each surgical

videotape and each novel videotape were seen an equal number

of times per group and per gender within each group.

Experimental paradigm. Table 8 presents the trial

sequence for the study, which is detailed below for each

experimental group. After the baseline recordings, the

sequence of trials began. All stimulus presentations were

separated by a variable length interval which averaged 1.5

minutes (range of 1 to 2 min) during which the subject

remained quietly seated, waiting for the next stimulus

presentation. The experimenter did not reenter the

experimental room until after the final (novel surgery)

trial, when he disconnected electrodes, debriefed (Appendix

I), and dismissed the subject.

The structure of all of the stimulus presentations

trials--both videotape and audiotape--was similar to that in

Study 1.1 Baseline HR and SCL were assessed during the 10 s

immediately prior to the onset of the stimulus, which was

presented for 60 s during which HR and SCL continued to be

recorded. For the four "test trials" (Exposures 1, 3, 5,

and 7 for the Video Only Group) and the Novel trial, the

subject's face was videotaped as he/she watched the surgery

videotape being presented. Faces were not videotaped during

audiotape presentations. Immediately upon stimulus offset,

BP was assessed, the affective ratings screen illuminated,

Table 8. Experimental Design for Study 2

Surgery Videotape Exposure

Video I 1 I I I I Novel
Only I la 2 1 3a 4 5a 6 I 7a ISurger
Group I I I I I I I Video

a Subjects' faces were videotaped for analysis of avoidance
and disgust.


IPrep ITest Prep ITest Prep ITest Prep ITest Novel
Group 1 1 2 2 3 3 4 4
G 0 __G 01 G 01 G OIV GO
Audiotapel IY P1 IY PI IY P IY PI Y P

and subjects rated the affect they experienced during the


Experimental Groups

Subjects were randomly assigned to one of three

experimental groups which varied in the type of

presentations they received prior to the Novel trial. For

the Novel trial, all subjects were presented a novel

surgical videotape, which was chosen randomly from among the

surgical videotapes not used for that subject; assignment of

novel stimulus was conducted to assure that all surgeries

were presented equally often within and between experimental

groups and for each gender. The details of stimulus

presentation for each of the three experimental groups were

as follows:

Video Only Group. The function of this group was to

evaluate the affect change during multiple repetitions of a

single surgery videotape followed by a novel surgery. Thus,

these 20 blood phobics were presented one particular surgery

scene seven times (listed in Table 8 by Exposure number),

followed by a single presentation of a novel surgery.2

Although these subjects wore the headphones just like other

subjects, they heard no audiotaped stimulus descriptions

during the study.

Two other groups were studied to compare the effects of

relevant and control preparatory verbal descriptions. For

these two groups, there were four preparation trials during

which an audiotaped description was presented repeatedly,

four test trials during which a particular surgery videotape

was presented repeatedly, followed by one presentation of a

novel surgery videotape.

Surgery Audiotape Group. During the four preparation

presentations, each subject in this group heard a relevant

and factual description of the upcoming surgery that was

presented during each of the four test trials. For each

subject, the audiotape was repeated four times, because the

same surgery videotape was repeated on the four test trials.

Neutral Audiotape Group. This group served as a

control for the Surgery Audiotape group. Each subject heard

a neutral audiotape description as "preparation" prior to

each test trial. Like the Surgery Audiotape subjects, these

subjects heard the same neutral audiotape for the four

preparations. On the four test trials they repeatedly

viewed one of the surgery videotapes, prior to viewing the

novel surgery videotape for the final trial. The inclusion

of this group permitted a controlled evaluation of the

effects of listening to a scene that lacked phobic content.


Three types of stimuli were used in this study, each of

which had five exemplars. First, in addition to the two

surgical videotapes employed in Study 1 ("Incision" and

"Tubes"), three additional 60 s surgical videotapes were

taken from Ost's film of thoracic operations. The three

additional video stimuli were the following: "Rib," which

showed a tool cleaning and removing a rib; "Heart," which

showed a beating heart being punctured with an instrument

and then sutured to stop bleeding; and "Sutures," which

showed the chest incision being sutured with needles and

thread. Blood and bodily deformation are depicted in all

five scenes. For each subject, one of these five surgery

videotapes was repeatedly presented, and one of the

remaining four was presented during the novel trial.

Sixty second audiotaped descriptions of each surgery

videotape were employed. In addition to the two audio

descriptions of "Incision" and "Tubes," three additional

audiotapes of 160 words narrated the three additional

surgery videotapes used in this study. The same female

voice recorded all audio stimuli in an effectively neutral

manner. These audiotapes were presented only to the Surgery

Audiotape subjects, and the surgery audiotape selected for

that subject was one that described the surgery videotape

that had been assigned to the subject.

The third category of stimuli were audiotaped verbal

descriptions of neutral, everyday activities. Five

audiotaped narratives, 60 s in duration, and 160 words long,

were recorded by the same female assistant who recorded the

surgical audiotape descriptions. The five neutral

descriptions included a person baking a cake, planting in

the garden, typing on a typewriter, paddling a canoe, and

flying a kite. These scenes were expected to be neutral in

affective content, yet all described human hand movement,

which was similar to the descriptions in the surgical

audiotapes. Only subjects in the Neutral Audiotape Group

heard these descriptions. Each neutral description was

presented an equal number of times with ordering of

presentation determined via a Latin square. (See Appendix E

for the transcripts of the five verbal and five neutral

descriptions.) In summary, since there were five surgery

and audiotape exemplars, four subjects of the 20 in each

experimental group received the same stimulus or stimulus


The practice neutral videotape stimulus which preceded

the experimental trials was a silent, 60 s videotape of a

scene from Jonathon Livingston Seagull which shows a seagull

in flight over mountains and the ocean.


The Mutilation Questionnaire and the Questionnaire Upon

Mental Imagery were already described in Study 1. The

additional questionnaire of interest in this study was the

Miller Behavioral Style Scale (MBSS). Miller (1980)

developed this instrument to assess a subject's preferred

coping style under stressful circumstances. The scale asks

the subject to imagine being in four stress-evoking

situations. Each scene is followed by eight statements

representing different ways of coping with the situation.

Four of the statements relate to confronting and seeking

information (a "monitoring" style), and the other four

statements indicate distracting and avoiding information

("blunting"). Separate monitoring and blunting scores are

obtained by summing the number of statements endorsed for

each coping style. The scale has good discriminant validity

(Miller, 1987a; Miller, Brody, & Summerton, 1988) and

predictive validity (Gard & Edwards, 1986; Miller, 1987a,

Phipps & Zinn, 1986, Watkins, Weaver, & Odegaard, 1986).

Dependent Measures

Three classes of dependent measures were assessed in

this study. Subjective measures included self-reported

displeasure, arousal, and lack of control, for which change

scores were calculated by subtracting the study baseline

value from the value obtain for each stimulus presentation.

Physiological measures of HR and SCL change scores were

calculated by subtracting the trial baseline mean (the 10 s

prior to stimulus onset) from the mean value obtained during

the stimulus. The two BP measures were derived by

subtracting the study baseline from the values obtain after

each stimulus.

Motor behavior measures of avoidance and maximum facial

disgust were coded by two independent raters from the

videotape of the subject. Avoidance was coded dichotomously

based on at least 1 s of avoidance. The 5-point rating of

maximum facial disgust was reduced to a dichotomous code to

provide a sufficient number of subjects for frequency

analyses; original scores of 0 or 1 were classified as "low"

disgust, and scores of 2, 3, or 4 were classified as "high"

disgust. Percent agreement for these two measures was

calculated between the two independent raters' dichotomous

classifications. For avoidance, percent agreement ranged

from 92% to 97% for the five trials. For maximum facial

disgust, percent agreement ranged from 93% to 98%. Thus,

acceptable interrater reliability was achieved.

This study used the same environment and apparatus as

in Study 1. A separate VPM control program was written to

assess physiology and present stimuli (Appendix J).

Three major issues were addressed in this study. In

the first section below, changes in affect to repeated

surgical exposures and to a novel surgery were evaluated

using data from the Video Only Group. In the second

section, the Surgery Audiotape and Neutral Audiotape groups

were compared to evaluate the effect of auditory preparation

on affect during repeated surgery presentations. In the

third section, the influence of imagery ability and coping

style on affect for those two groups of subjects who

received auditory preparations was examined.

Effects of Repeated Visual Exposure

The 20 subjects of the Video Only Group viewed one

surgery videotape seven times prior to viewing a novel

surgery. The dependent measures of displeasure, arousal,

lack of control, SCL, HR, SBP, and DBP were assessed during

each of these eight exposures. These measures were analyzed

with a repeated-measures ANOVA in which Trial (eight levels)

was the within-subject effect. The use of Greenhouse-

Geisser corrections resulted in fractional degrees of

freedom. Planned contrasts between exposures were conducted

to more thoroughly evaluate affect change.

Self-report measures. Table 9 presents displeasure,

arousal, and lack of control change scores from baseline

over the seven exposure trials and the Novel Exposure for

Video Only subjects. As indicated by significant Trial

effects and planned contrasts between exposures, all three

variables showed a significant reduction in negative affect

from Exposure 1 to Exposure 7, and a significant and

immediate return of negative affect to the Novel Exposure to

levels not significantly different from those reported

during Exposure 1 (all R > .42). At Exposure 7, arousal and

lack of control did not differ from their study baselines (R

> .65), although displeasure remained above its baseline

value, (Trial effect: displeasure, E(3.2, 60.3) = 8.61, R <

.0001; arousal, (2.8, 52.6) = 10.34, R < .0001; lack of

control, E(3.0, 56.7) = 7.76, E < .0002.

Physiological measures. Figure 2 presents the SCL and

HR change data for the Video Only subjects on the same graph

for easy comparison. The trend in SCL and HR parallels that

of the self-report data, although only for SCL was there a

significant Trial effect, (2.6, 49.1) = 4.28, R = .012.

The SCL habituated rapidly after an initial elevation during

Exposure 1, dropped significantly to its lowest value during

Exposure 5 (which was not significantly different from

baseline, R > .44), and then increased somewhat (although

nonsignificantly) during the Novel surgery. The SCL during

Table 9. Changes in Self-Report and Motor Measures Across
Repeated Surgery Videotape Exposures and a Novel
Surgery for the Video Only Group



1 2 3 4 5 6 7 Novel

Displeasure 6.5 5.9
(4.0) (3.6)


5.9 4.4
(8.2) (8.3)

Control Lack 5.3 4.7
(5.8) (6.1)

(Yes/No) 8/12
(High/Low) 10/10

4.9 4.6
(3.6) (3.7)

3.5 2.2
(8.2) (8.0)

3.4 1.9
(6.3) (6.6)










3.7 7.2
(4.5) (8.2)

0.4 6.4
(7.9) (7.5)

1.2 0.7 5.7
(6.7) (6.9) (7.1)

10/10 9/11

5/15 9/11

a Means (and standard deviations) are presented for the
self-report measures; frequency data are presented for the
motor measures.

5 +



4 +


3 +


2 +


1 +






+ 0.6

+ 0.4

+ 0.2

+ 0.0


+ 1.4

S.....S Skin Conductance

H--- H Heart Rate + 1.2

+ 1.0

1 + 0.8

1 2 3 4 5 6 7 Novel


Figure 2. Changes in Heart Rate and Skin Conductance Level
Across Seven Surgery Videotape Exposures and the
Novel Surgery Exposure for the Video Only Group






the novel surgery did not differ from that during Exposure 1

(p > .12). Neither SBP nor DBP change scores showed a
statistically reliable trend across exposures.

Motor behavior measures. The motor measures of

avoidance and facial disgust were assessed during only

Exposures 1, 3, 5, 7, and the Novel Exposure and are

presented in Table 9. These dichotomous measures were

analyzed via chi-squares using McNemar's (1969)

recommendation for within-subject analyses. As can be seen,

avoidance behavior remained at a relatively constant level

across the four exposures and the novel surgery, except for

a nonsignificant drop in avoidance during the Exposure 5.

For facial disgust, half of the Video Only subjects showed

"high" disgust during Exposure 1; this frequency decreased

in subsequent exposures and the difference reached

statistical significance at Exposure 7, X2(1) = 5.0, R <

.05. Additionally, significantly more subjects increased

than decreased disgust from Exposure 7 to the Novel Surgery,

X2(1) = 4.00, R < .05. Disgust during the novel surgery and
Exposure 1 did not differ.

Effects of Auditory Preparation

The effects of auditory preparation were examined by

comparing the affect of the relevantly-prepared Surgery

Audiotape Group with the control Neutral Audiotape Group at

the four videotape "test trials" and the novel surgery.3

Data analyses used mixed-model repeated measures ANOVAs in

which Group (Surgery Audiotape or Neutral Audiotape) was the

between-subjects effect and Trial (five levels) was the

within-subject effect. Greenhouse-Geisser corrections were


Self-report measures. Figure 3 shows the displeasure

change during each of the four test trials and the novel

trial for Surgery Audiotape and Neutral Audiotape subjects.

For all three self-report measures, negative affect

decreased significantly between Test Trials 1 and 4 and

increased significantly from Test Trial 4 to the Novel

Surgery (all R < .007). Group differences were of greater

interest, however. As Figure 3 suggests, Surgery Audiotape

subjects reported less displeasure than the Neutral

Audiotape subjects across trials, Group main effect, F(1,

38) = 4.40, R = .042. The Trial X Group interaction was not

significant, indicating that the group difference in

displeasure did not change across trials. Lack of control

followed a similar trend across trials in that the sample

mean of the Surgery Audiotape Group suggested they had less

lack of control than the Neutral Audiotape Group; however,

this Group main effect failed to reach significance, R =

.16. Self-reported arousal was similar for the two groups

across all trials.

Physiological measures. Figure 4 shows the change in

SCL across the four test trials and novel trial for the two

groups. There was a significant decrease in SCL over the

four test trials and increase during the Novel Trial, (R <

.001). More importantly, Figure 4 suggests that the SCL of

10 +


E 7+


SA.....SA Surgery Audiotape

NA---- NA Neutral Audiotape


SA.. .SA



---+----------+----------+------------ ---- -

Test 1

Test 2

Test 3


Test 4

Figure 3.

Self-Reported Displeasure Change Across Four
Surgery Videotape Test Trials and the Novel
Surgery Trial for the Surgery and Neutral
Audiotape Groups



S 1.2 + NA
I SA.....SA Surgery Audiotape
NA---NA Neutral Audiotape
C 1.0 +
C 0.8 +
E 0.6 +

N 0.4 +

0.2 + .NA
C "
R .
M 0.0+ .
N, *SA .. ..... SA
-0.2 +

---- ---------+---------+--------------------
Test 1 Test 2 Test 3 Test 4 Novel


Figure 4. Skin Conductance Level (60 s Mean) Change Across
Four Surgery Videotape Test Trials and the Novel
Surgery Trial for the Surgery and Neutral
Audiotape Groups

the Surgery Audiotape Group was less than that of the

Neutral Audiotape Group; however, the Group main effect only

neared significance, F(1, 38) = 2.73, R = .10. Nonetheless,

simple effects analysis revealed that at Test Trial 1, the

SCL of the prepared subjects was less than that of the

control subjects, R < .03. There was no Trial X Group

interaction. The HR, SBP, and DBP changes did not differ

significantly across trials or between groups.

Motor behavior measures. Table 10 presents avoidance

and facial disgust frequency data for the Surgery Audiotape

and Neutral Audiotape Groups. Consistent with the findings

for displeasure and SCL, these motor measures indicated that

the prepared subjects had less negative affect during the

surgery videotapes than the control subjects.

Avoidance did not change significantly for either group

across the five trials. "High" disgust for the Neutral

Audiotape subjects steadily declined from Test Trial 1 to

Test Trial 4, X2(1) = 7.0, R < .01; the fairly low disgust

displayed by the Surgery Audiotape subjects did not decrease

across test trials. Both groups had significantly more

subjects increase than decrease disgust from Test Trial 4 to

the Novel Trial, (Surgery Audiotape, X2(1) = 4.00, R < .05;

Neutral Audiotape, X2(1) = 7.00, R < .01).

Next, differences at each trial were examined. The

Surgery Audiotape Group tended to avoid less frequently than

the Neutral Audiotape Group at Test Trials 1 and 4, (X2(1) =

2.50 and 7.06, R = .114 and .058, respectively), and they

Table 10.

Frequency of Avoidance and Facial Disgust Across
Four Test Trials and the Novel Surgery for
Surgery Audiotape and Neutral Audiotape Groups

Behavioral Measure

Avoidance (Yes/No)

Surgery Audiotape
Neutral Audiotape



Test 1 Test 2 Test 3 Test 4






Surgery Audiotape
Neutral Audiotape








avoided significantly less often than the control subjects

during Test Trial 2 and the Novel Trial, (X2(1) = 7.62 and

7.06, R = .006 and .008, respectively). The Surgery

Audiotape Group had significantly fewer "high" disgust

subjects than the Neutral Audiotape Group during Test Trials

1, 2, 3, and the Novel Trial, (X2(1) = 5.23, 8.53, 4.33, and

3.75; R = .022, .003, .037, and .053, respectively).

Effects of Personality Variables

Two personality variables, imagery ability and coping

style, were examined for their influence on affect of

subjects receiving preparation--the Surgery Audiotape and

Neutral Audiotape Groups. Analyses used repeated-measures

ANOVAs with Group and the personality measure as between-

subjects effects; personality measures were left continuous

for the ANOVA, but were dichotomized via a median split for

chi-square analyses of their relationship with avoidance and

disgust and for graphic presentation.

Mental imagery ability (QMI). Imagery ability was

hypothesized to influence affect during the auditory

preparation trials; thus, affect during Preparations 1 and 2

was analyzed. Only self-reported lack of control was

related significantly to imagery ability. Examination of

the plotted data revealed an effect of little interest: the

lack of control of good imaging Neutral Audiotape subjects,

(but not Surgery Audiotape subjects) was lower during both

of the neutral audiotape descriptions, (QMI X Group, F(1,

36), E = .033).

Coping style (MBSS). The MBSS yielded both Monitoring

and Blunting scores which were not correlated in this sample

of 40 subjects, (r(38) = -.16, 2 = .31); therefore, each

variable was examined separately. Coping style was expected

to relate to emotional responding during the actual

presentation of the surgery stimuli. Thus, repeated-

measures ANOVAs and chi-squares examined each variable

during Test Trials 1 and 2 and again during Test Trial 4 and

the Novel Surgery.

The Monitor variable was related to only one dependent

measure; all five of the Surgery Audiotape subjects who

showed "high" disgust to the Novel Surgery were low in

Monitoring, X2(l) = 6.67, E = .01.

The Blunting variable, however, showed several

interesting relationships with various dependent measures.

The self-report measures were related to Blunting during

Test Trials 1 and 2, whereas several of the physiological

measures were related to Blunting during Test Trial 4 and

the Novel Surgery. Across Test Trials 1 and 2, high

blunting subjects reported a rise in displeasure, whereas

low blunting subjects reported an decrease in displeasure,

regardless of Group, (Trial X Blunting interaction, f(1, 36)

= 5.10, R = .03). The analyses of arousal and lack of

control across Test Trials 1 and 2 were consistent with

displeasure and even more revealing in that the groups

differed. The interactions for both dependent measures were

interpreted similarly; therefore, only arousal change scores

for the two groups as a function of blunting are presented

in Figure 5. As the figure reveals, Surgery Audiotape high

blunters reported increased arousal and lack of control

across the two surgery presentations, whereas low blunting

Surgery Audiotape subjects reported a reduction in arousal

and lack of control. Neutral Audiotape subjects, however,

showed little change in arousal or control across these

surgeries as a function of blunting; indeed, Neutral

Audiotape high blunters reported somewhat greater arousal

and lack of control during these presentations than low

blunters (Trial X Group X Blunting interactions for arousal,

F(l, 36) = 8.78, p = .005; and lack of control, F(1, 36) =

13.44, R = .0008).

Blunting was related also to affect during Test Trial 4

and the Novel Surgery. For SCL, collapsing across both

trials, the Surgery Audiotape high blunters tended to have a

higher SCL than Surgery Audiotape low blunters, whereas

Neutral Audiotape high blunters had a lower SCL than Neutral

Audiotape low blunters, (Group X Blunting, F(1, 36) = 3.75,

R = .06). The two BP measures specified these effects for

each trial, and since the interpretation of the interactions

is similar, only SBP data for both groups over both surgery

presentations as a function of Blunting is presented in

Figure 6. Consistent with SCL, SBP and DBP increased from

Test Trial 4 to the Novel Trial for high blunting Surgery

Audiotape and low blunting Neutral Audiotape subjects,

whereas SBP and DBP showed little change across surgery

Surgery Audiotape
13 +

12 +

11 +

10 +









Figure 5.

Neutral Audiotape



Low Blunting

High Blunting


L. ..........L


Arousal Change Across Test Trials 1 and 2 for the
Surgery Audiotape and the Neutral Audiotape
Groups as a Function of Blunting







Test 4


Neutral Audiotape

Surgery Audiotape
8 +

6 + H

4 +

2 +


0 +


-2 + L

-4 + H


Test 4



Figure 6. Systolic Blood Pressure Change Across Test Trial
4 and the Novel Trial for the Surgery Audiotape
and the Neutral Audiotape Groups as a Function of

L.....L Low Blunting

H----H High Blunting

presentations for low blunting Surgery Audiotape and high

blunting Neutral Audiotape subjects, (Trial X Group X

Blunting interactions for SBP, F(1, 36) = 8.35, R = .0065;

and DBP, F(1, 36) = 5.12, p = .030). Finally, findings for

self-reported arousal were not completely consistent with

the physiological data. Both Neutral Audiotape and Surgery

Audiotape low blunters reported a greater increase in

arousal from the Test Trial 4 to the Novel Trial than did

the high blunters, (Trial X Group X Blunter interaction,

F(l, 36) = 4.86, p = .034).

Affect Change to Repeated Surgery Exposures

A fundamental question regarding emotional change is

whether blood phobics become less uncomfortable or anxious

as a phobic stimulus is presented repeatedly. This study

indicated that such change does occur. Indeed, in a sample

of 20 phobics who viewed one surgery stimulus repeatedly for

seven trials, there was a significant decline on most

measures of negative affect across repetitions. This

finding supports a basic premise of emotional functioning--

exposure leads to reductions in negative affect (Foa &

Kozak, 1986).
Similar findings were reported by Hare et al. (1971)

who demonstrated habituation of arousal with multiple

repetitions of mutilation scenes. Hare and colleagues

studied normals using slides; the current study extends

their findings to blood phobics viewing videotapes.

Generalization of Affect Reduction to Novel Phobic Stimuli

A second question of fundamental importance concerns

the extent to which the attenuation of emotion which

occurred during repeated presentations generalized to new

blood-related stimuli. The results of the current study

indicated that, although the aversive qualities of a

particular blood-related stimulus attenuated fairly quickly

to repetition, very little or no generalization occurred.

Indeed, when a novel surgery scene was presented, there was

a substantial return of negative affect to levels not

different from those observed during the first presentation

of the surgery scene. Using a different paradigm, Hare et

al. (1971) found that presenting different mutilation slides

interfered with habituation. Limited generalization to a

novel stimulus might be viewed as secondary to affective

network changes in only those specific stimulus properties

shown in the repeated surgery scene, rather than to a

elaborated network of "blood-related stimuli." A further

test of generalization might be to repeatedly present the

"novel" stimulus. If the rate of habituation to it is

faster than to the original stimulus, evidence for some

generalization of affect reduction would be adduced.

Effects of Preparation

A second portion of this study examined the effects of

two types of preparation for upcoming surgery videotape

exposures. The prepared group of blood phobics heard an

accurate, effectively neutral description of the surgery

prior to each of four viewings. Their affect during the

four repeated exposures and during a novel stimulus was

compared to that of a control group of subjects who heard a

neutral, unrelated audiotaped description before each


Generally, the relevant preparation produced only

modest reductions in negative affect during exposure,

compared with the control preparation. The relevantly

prepared group reported significantly less displeasure and

showed less facial disgust and avoidance than did control

subjects. The lack of control and SCL change showed similar

although less reliable trends. Preparation effects tended

to be somewhat more evident during the first surgery

exposure than later exposures. Thus, although the effect

does not appear robust, there is tentative support for the

hypothesis that preparing blood phobics to view a surgery by

providing them a description modestly reduces negative

affect in comparison to an irrelevant preparation.

Admittedly, however, group differences were limited

with respect to the number of variables differentiating

conditions and the magnitude of the effects. One important

potential reason for the limited effects is that the current

paradigm simply did not permit large effects. In this

study, the experimental manipulation was a variation in

preparation for blood phobics viewing surgeries; one might

expect rather small or even absent effects from this

manipulation, especially in light of the limited effects

found in Study 1, where powerful experimental conditions

were created phobicss versus nonphobics, aversive versus

neutral material).

The content of the preparatory audiotapes may further

account for limited effects. The affect of the Surgery

Audiotape group during the first audiotape presentation was

minimally negative; only two subjective measures indicated

increased arousal. The narrative of each audiotape was

effectively neutral and purely descriptive of the surgery;

it contained no references to the observed patient's

experience of the surgery nor to the subject's possible

reaction during viewing. The empirical literature on

preparation for stressful procedures has demonstrated that

verbal preparations which include not only descriptions of

upcoming procedures but also of the sensations that the

patient or listener might experience, (i.e., sensory

information) often result in less anxiety during and after

the stressful event (Anderson, & Masur, 1983). A different

domain of empirical inquiry--affective imagery--suggests

that verbal scripts for emotional imagery evoke increased

affect when they include response propositions, or

descriptions of the imaging person's affective reactions in

the imaged scene (Lang, Kozak, Miller, Levin, & McLean,

1980; Lang, Levin, Miller, & Kozak, 1983). Additionally,

the current subjects were instructed only to listen to the

description, whereas other research has demonstrated that

instructions to vividly imagine oneself in the scene elicits

greater affect during imagery (Lang, 1979). Thus, the

preparatory capacity of the audiotaped descriptions might

have been enhanced by modifying the instructions and

incorporating response propositions.

Individual Difference Variables

Finally, another reason for limited group preparation

effects is the variability in affect accounted for by

individual differences, especially in coping style. Imagery

ability was assessed also, but it was found to have little

relation to affect during the verbally presented preparation

descriptions, where imagery effects might have been expected

to occur.

The effects of coping style appear to be more robust,

however. The monitoring style was limited in its relation

to affect during the surgery videotapes. Its one

relationship was consistent with the effects of the blunting

style and is described below. The blunting coping style was

related to increases or decreases in self-reported affect

and physiology from the first to the second surgery

videotape presentation and the fourth surgery to the novel

presentation. Moreover, blunting coping style interacted

with the type of preparation given to subjects, yielding

different relationships for the two preparation groups.

Among the relevantly prepared subjects, blunters

reported lower levels of negative affect to the first

surgery presentation, but their negative affect increased

during the second surgery presentation. Subjects who

typically avoid blunting reported greater negative affect on

the first presentation, but their anxiety decreased to the

second surgery. The unprepared subjects did not show this

effect, but tended to display the opposite relation.

Similar effects occurred during the presentation of a

novel surgery as indexed by physiological measures.

Prepared subjects who blunt showed an increase in

physiological arousal to the novel surgery compared with

prepared subjects who do not blunt. Also, prepared subjects

who were low on monitoring (typically considered as similar

to blunting) had the greatest facial disgust to the novel

surgery. Again, unprepared subjects did not show this

pattern of results. In the transition from the repeated

surgery to the novel surgery, unprepared control subjects

who do not blunt had increased physiological arousal.

The findings with the prepared subjects are consistent

with those of Shipley and colleagues (1978, 1979) who found

that repressors (like blunters in the current study) became

more anxious (indexed by tachycardia) during a second

presentation of a surgery preparation videotape, whereas

sensitizers were more anxious during the first presentation

and became less so during a second viewing. This study

extends their findings by suggesting that when a novel

surgery is shown later in the repetition sequence, low

blunters (like sensitizers) continue to show reduced affect,

but blunters (like repressors) are increasingly disturbed

and unable to blunt as successfully as they did during the

initial presentation.

The different relationship of blunting with affect for

the preparation and control groups is interesting. It is

possible that the surgical descriptions for the prepared

subjects permitted almost continuous activation of their

affective networks, with the result that the subjects'

preferred manner of coping with aversive stimuli had

predictable effects over exposures. However, providing

other phobics irrelevant, potentially distracting

descriptions prior to each viewing might have precluded

continuous activation of the emotion, with the result that

low blunters were unable to decrease their arousal over

presentations, and high blunters successfully maintained

affective distance. Naturally, these explanations are

speculative, and since the construct of coping style is

itself poorly understood, it is quite difficult to enlighten

the complicated interaction of coping style with variations

in exposure. Further theory and research which addresses

the mechanism by which coping style influences emotional

network activation is needed.

Methodological Issues

Study 2's methodology had several problems which

hindered clear interpretations of the results and probably

reduced the size of observed effects. First, the study used

five different surgery scenes rather than a single scene or

two, as in Study 1. It was hoped that multiple surgeries

would increase the generalizability of the findings to a

larger population of blood-related stimuli. Although no

experimental confound occurred (each surgery was used

equally often), the increase in response variation

attributable to the use of multiple surgeries probably

resulted in increased between-subjects variance and a

concomitant decrease in statistical power. Additionally,

too many surgeries were employed to permit adequate

statistical comparisons of their affect-eliciting power.

Ideally, the same two surgeries used in Study 1 should have

been employed in Study 2 to more clearly replicate and

extend the findings of Study 1.

Second, avoidance and facial disgust should have been

recorded during verbally presented descriptions in both

studies. Bioinformational theory predicts that various

stimulus modalities reliably activate affective networks,

resulting in measurable efferent outflow (Lang, 1979).

Thus, future research should include facial affect

assessments during verbal stimulus trials. A more important

experimental concern is that no baseline avoidance or facial

affect measures were recorded; thus, it is possible that the

observed differences between groups were attributable to

preexisting differential tendencies to avoid or show

disgust. An acceptable baseline might be obtained by

presenting a stimulus that lacks the phobic content under

study, but is hypothesized to be equally bothersome to all

experimental conditions, such as a snake or height stimulus

in studies of blood phobics and controls.

1 For the first 30 subjects only, BP was assessed during the
20 s prior to each stimulus onset, with the hope of using
this measure as a baseline. This procedure was discontinued
when it was discovered that the frequent cuff inflation was
excessively uncomfortable for subjects and may have cued
them as to the timing of stimulus onset. Discontinuation of
this assessment occurred after an equal number of subjects
(balanced for gender) from each group had completed
participation and after each surgical stimulus had been
presented an equal number of times. Therefore, no
experimental confound occurred. The prestimulus BP data
collected on the 30 subjects were not analyzed.
2 Prior to Exposure 1, Video Only subjects were not
presented any stimulus while the other groups were hearing
their first preparation audiotape. It was originally
intended that the lack of this stimulus presentation would
permit the Video Only subjects to serve as a "waiting, no
intervention control" for the auditory preparations
presented to the other two groups.

3 Since the Video Only Group received no preparation for the
first exposure, their affect during Exposure 1 was compared
with that of the other two groups during Test Trial 1. The
Video Only Group did not differ from the Neutral Audiotape
Group on any measures, and only differed from the Surgery
Audiotape Group in having more frequent avoidance and "high"

4 Affect differences between the Surgery and Neutral
Audiotape Groups during the initial audiotape preparation
per se were examined to determine whether hearing the
surgery description led to greater negative affect than
hearing the neutral, control description. As expected,
Surgery Audiotape subjects reported more displeasure and
arousal than Neutral Audiotape subjects (both E < .0002).
Lack of control and SCL followed this same pattern, but
neither reached statistical significance. Neither HR nor
the two BP measures differed between groups during this
preparation audiotape.


An aversive negative reaction to blood, injury, or

bodily deformation is a fairly common phenomenon

traditionally termed "blood phobia" and classified as a

simple phobia. Research has enlightened possible etiologies

(Kleinknecht, 1987; Ost & Hugdahl, 1985), examined subjects'

cognitions during exposure to blood-related stimuli

(Kaloupek & Stoupakis, 1985; Kaloupek, Scott, & Khatami,

1985), and developed exposure-based techniques with

modifications to prevent fainting (Ost & Sterner, 1987).

Generally, however, it has been less researched than other

simple phobias, perhaps because most authors view it as

simply another phobia (Marks, 1988). Yet its unique

features hinder straightforward extrapolation from the

extensive theory and empirical literature on simple phobias.

A Comparison of Blood Phobia and Other Phobias

The literature on blood phobia has consistently used

the term "fear" to describe the affect of phobics vis-a-vis

blood-related stimuli. For example, the Mutilation

Questionnaire is considered to measure respondent's "fear"

of mutilation stimuli (Klorman et al., 1974). One wonders,

however, what is it specifically that is feared? It is

unlikely that the external blood-related stimulus itself is

feared, for what damage or harm can it accomplish? Beck and

Emery (1985) noted that blood phobics do not report fear

during actual confrontation with the stimulus, rather, they

feel queazy, disgusted, or squeamish. Additionally, the

facial expression exhibited by the blood phobics in the

current study was routinely one of disgust rather than fear.

Thus, it appears that the subjective experience of blood

phobics is complicated, possibly encompassing fear prior to

or early during exposure, but that another affect such as

disgust is dominant during exposure. More detailed study of
the subjective emotional experience of blood phobics both

before and during exposure is worthwhile.

The psychophysiology of blood phobia is thought to be

unique among phobias. In other simple phobias, exposure to

the phobic stimulus results in a prolonged sympathetic

response with classic markers of fear and anxiety such as

tachycardia, hypertension, and increased sweating and

respiration rate. Yet blood phobia is considered to have a

biphasic response pattern of sympathetic activation followed

by parasympathetic activity which, if exposure is continued,

leads to fainting. Research indicates that fainting is

uniquely associated with blood phobia and not with other

phobias (Connolly, Hallam, & Marks, 1976), and many

researchers suggest that most blood phobics faint (Marks,

1988). Yet of the 84 blood phobics participating in both

studies, none fainted nor appeared to near faint during

exposure to the surgeries.

There are several potential reasons for the

discrepancies between the current studies and past research.

It is possible that these subjects were not sufficiently

phobic, and that more severe blood phobics would have

fainted. Several clinical studies have found about a 70%

prevalence of fainting in the histories of blood phobics

presenting for treatment (Ost, Sterner, & Lindahl, 1984;

Connolly et al., 1976; Thyer et al., 1985). However,

generalizations from clinic patients to all blood phobics

may not be appropriate. It is also possible that the

stimuli used in the current studies were insufficiently

aversive either due to the content or duration. In vivo

exposure to an operation or undergoing venipuncture and

blood donation probably are more aversive and likely to

elicit fainting (Graham et al., 1961). Regarding stimulus

duration, brief exposures to mutilation stimuli resulted in

tachycardia but not significant bradycardia in Study 1 and

in studies by Klorman et al. (1975, 1977). Longer duration

presentations (Ost et al., 1984; Steptoe & Wardle, 1988)

more frequently find parasympathetic activity and fainting.

Thus, longer stimulus presentations may permit the

occurrence of the parasympathetic portion of the biphasic

reaction, whereas shorter presentations permit only

sympathetic activity.

It is possible, however, that fainters constitute a

distinct group, only partially overlapping with blood

phobics. Kleinknecht and Lenz (1989) recently found that

the population of people who report fear to blood stimuli

can be subdivided into fainters and nonfainters, and that

some fainters report little or no fear of blood stimuli.

This apparently accounts for Kleinknecht's (1988a) earlier

finding of only a modest correlation (r = .30) between MQ

scores and a history of fainting to blood-injury stimuli.

Thus, aversion or fear of blood-related stimuli appears to

be less tightly associated with fainting than an examination

of clinic blood phobics and blood donation fainters would

lead one to believe. Alternatively, Kleinknecht's blood

phobics and those in the current studies were college

students and younger than patients in the above clinical

studies. It is possible that these subjects were unsure

whether or not they would faint, because they too rarely

have encountered blood-related stimuli or because they

reliably escape or avoid. Future research might find that

if such people were to continue exposure to a sufficiently

intense blood-related stimulus, perhaps most or all would

faint. Thus, an important area of research is an evaluation

of the extent of fainting among blood phobics, those

situations in which fainting occurs, and the differences

between fainters and nonfainters.

Behavioral avoidance testing to assess the motoric fear

response is commonplace in studies of simple phobias. With

blood phobia, however, only Ost's research team has used a

behavioral measure: duration of viewing a prolonged surgery

film. Blood phobia may differ from other phobias in that